Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "Bulletin of the British Museum (Natural History), Geology"

See other formats


Sgfff 


BULLETIN  OF 

THE  BRITISH  MUSEUM 

(NATURAL  HISTORY) 


GEOLOGY 
VOL.  XV 
1967-1968 


TRUSTEES  OF 

THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 

LONDON:  1970 


DATES  OF  PUBLICATION  OF  THE  PARTS 


No.   i.  26th  May 

•      1967 

No.   2.   13th  June 

•      1967 

No.   3.   24th  November 

.      1967 

No.  4.   12th  January- 

.      1968 

No.   5.   19th  March     . 

.      1968 

No.  6.   14th  May 

.      1968 

PRINTED  IN  GREAT  BRITAIN 
BY  ALDEN  &  MOWBRAY  LTD 
AT  THE  ALDEN  PRESS,  OXFORD 


CONTENTS 

GEOLOGY  VOLUME  XV 

No.  i.     The  palaeontology  and  stratigraphy  of  the  lower  part  of  the  Upper 

Kimmeridge  Clay  of  Dorset.     J.  C.  W.  Cope  i 

No.  2.     The    correlation    and    trilobite    fauna    of    the    Bedinan    Formation 

(Ordovician)  in  south-eastern  Turkey.     W.  T.  Dean  8i 

No.  3.     Burrows  and  surface  traces  from  the  Lower  Chalk  of  southern  England. 

W.  J.  Kennedy  125 

No.  4.     A  new  Temp  sky  a  from  Kent.     M.  E.  J.  Chandler  169 

No.  5.     Colonial  Phillipsastraeidae  from  the  Devonian  of  south-east  Devon, 

England.     C.  T.  Scrutton  181 

No.  6.     Some  Strophomenacean  brachiopods  from  the  British  Lower  Silurian. 

L.  R.  M.  Cocks  283 

Index  to  Volume  XV  325 


THE  PALAEONTOLOGY  AND 
STRATIGRAPHY  OF  THE  LOWER 
PART  OF  THE  UPPER  KIMMERIDGE 
CLAY  OF  DORSET 


J.  C.  W.  COPE 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  i 

LONDON:   1967 


2  5  MAY  196 

THE  PALAEONTOLOGY  AND   STRATIGRAPHY  W^  ^ 
OF  THE  LOWER  PART  OF  THE   UPPER 
KIMMERIDGE   CLAY  OF  DORSET 


BY 


J.  C.  W.  COPE      , 

Department  of  Geology,  University  College,  Swansea 


Pp.  1-79;  33  Plates;  12   Text-figures 


BULLETIN  OF 
THE  BRITISH  MUSEUM   (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  1 

LONDON:   1967 


THE    BULLETIN    OF    THE    BRITISH    MUSEUM 

(NATURAL  HISTORY),  instituted  in  1949,  is 
issued  in  five  series  corresponding  to  the  Departments 
of  the  Museum,  and  an  Historical  series. 

Parts  will  appear  at  irregidar  intervals  as  they  become 
ready.  Volumes  will  contain  about  three  or  four 
hundred  pages,  and  will  not  necessarily  be  completed 
within    one    calendar   year. 

In  1965  a  separate  supplementary  series  of  longer 
papers  was  instituted,  numbered  serially  for  each 
Department. 

This  paper  is  Vol.  15,  No.  1  of  the  Geological 
(Palaeontological)  series.  The  abbreviated  titles  of 
periodicals  cited  follow  those  of  the  World  List  of 
Scientific  Periodicals. 


World  List  abbreviation 
Bull.  Br.  Mus.  not.  Hist.  (Geol.) 


Trustees  of  the  British  Museum  (Natural  History)  1967 


TRUSTEES    OF 
THE    BRITISH    MUSEUM    (NATURAL    HISTORY) 

Issued  26  May,  1967  Price  £4  15s. 


THE  PALAEONTOLOGY  AND   STRATIGRAPHY 

OF  THE  LOWER  PART  OF  THE   UPPER 

KIMMERIDGE   CLAY  OF  DORSET 

By  J.  C.  W.  COPE 


CONTENTS 


I.    Introduction 
II.    The  Kimmeridge  section 

III.  The  non-ammonite  fauna 

IV.  The  ammonite  fauna 

(a)  Introduction 

(b)  Systematic  descriptions 

Genus  Gravesia  Salfeld 

Gravesia  gigas  (Zieten) 

G.  cf.  gravesiana  (d'Orbigny) 
Genus  Pectinatites  Buckman 

Sexual  Dimorphism 

Interpretation  of  Pectinatites 

The  Origins  of  Pectinatites 
Subgenus  Arkellites  nov. 

Pectinatites  {Arkellites)  primitivus  sp 

P.  {A.)  cuddlensis  sp.  nov. 

P.  (A.)  damoni  sp.  nov. 

P.  (A.)  hudlestoni  sp.  nov. 
Subgenus  Virgatosphinctoides  Neaverson 

Pectinatites  {Virgatosphinctoides)  elegans  sp. 

P.  (V.)  elegans  corniger  subsp.  nov 

P.  {V.)  scitulus  sp.  nov. 

P.  (V.)  decorosus  sp.  nov.  . 

P.  (V.)  major  sp.  nov. 

P.  (V.)  clavelli  sp.  nov. 

P.  (V.)  smedmorensis  sp.  nov. 

P.  (V.)  laticostatus  sp.  nov. 

P.  (V.)  grandis  (Neaverson) 

P.  (V.)  grandis  acceleratus  subsp.  nov 

P.  (V.)  woodwardi  (Neaverson)  . 

P.  (V.)  wheatleyensis  (Neaverson) 

P.  (V.)  wheatleyensis  minor  subsp.  nov 

P.  (V.)  wheatleyensis  delicatulus  (Neaverson 

P.  (V.)  pseudoscruposus  (Spath) 

P.  (V.)  reisiformis  sp.  nov. 

P.  (V.)  reisiformis  densicostatus  subsp.  nov. 

P.  (V.)  abbreviatus  sp.  nov. 

P.  (V.)  donovani  sp.  nov.  . 

P.  (V.)  magnimasculus  sp.  nov 

P.  (V.)  encombensis  sp.  nov. 


Page 

4 

5 

10 

ii 

ii 

12 
12 
12 

14 
14 
15 
18 
22 
24 
24 
26 
28 
29 
30 
3i 
33 
34 
36 
37 
39 
4i 
43 
44 
45 
45 
46 

47 
48 

49 
50 
53 

54 
55 
56 
57 


GEOL.  15,  I. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET 


Subgenus  Pectinatites  Buckman     . 

Pectinatites  (Pectinatites)  inconsuetus  sp.  nov. 
P.  (P.)  eastlecottensis  (Salfeld)     . 
P.  (P.)  cf.  groenlandicus  (Spath) 
P.  (P.)  cornutifer  (Buckman) 
P.  (P.)  naso  (Buckman)     .... 
Genus  Pavlovia  Ilovaisky,  subgenus  Paravirgatites  Buckman 
Pavlovia  (Paravivgatites)  cf.  paravirgatus  (Buckman) 
V.    Evolution  of  the  ammonites 
VI.   The  ammonite  zones 

Pectinatites  ( Virgatosph.inctoid.es) 
P.  (V.)  scitulus  Zone 
P.  (V.)  ivheatleyensis  Zone 
P.  (Arkellites)  hudlestoni  Zone 
P.  (Pectinatites)  pectinatus  Zone 
VII.    Correlations  .... 

(a)  Great  Britain 

(b)  The  Boulonnais 

(c)  East  Greenland  (Milne  Land)   . 

(d)  Southern  Germany  (Franconia) 

(e)  Russia  (Basin  of  the  Ural  and  Ilek  Rivers) 
VIII.    References    ....... 


elegans  Zone 


Page 

59 
59 
60 
61 
62 
62 
63 
63 
64 
66 
66 
68 
69 
69 
70 
70 
70 
72 
73 
73 
75 
76 


SYNOPSIS 

Re-examination  of  the  lower  part  of  the  Upper  Kimmeridge  Clay  of  the  type  section  at  Kim- 
meridge,  Dorset,  has  involved  detailed  collecting  and  re-measurement  of  the  succession.  The 
ammonite  fauna  of  some  465  ft.  of  these  beds  is  described. 

The  ammonites  belong  to  three  subfamilies  ;  three  genera  are  represented,  and  descriptions 
are  given  of  three  subgenera  (one  new)  ;  twenty-eight  species  (seventeen  new)  ;  and  five  sub- 
species (four  of  them  new). 

Sexual  dimorphism  is  recognized  in  one  genus,  and  is  believed  to  be  of  a  unique  type. 

A  revised  zonal  scheme  is  proposed,  and  correlations  are  suggested  with  Northern  France, 
Greenland,  and  other  areas  of  Britain. 

Comparisons  are  made  with  the  ammonite  fauna  of  other  extra-British  areas.  The  fauna  of 
the  Lower  Tithonian  rocks  of  Germany  is  shown  to  be  quite  distinct  from  that  of  the  Upper 
Kimmeridgian  of  Britain,  such  similarities  as  exist  being  entirely  due  to  homeomorphy. 


I.     INTRODUCTION 

Existing  knowledge  of  the  Upper  Kimmeridgian  ammonite  faunas  of  Britain  is 
very  imperfect.  The  research  work  which  forms  the  basis  of  this  account  is  the  first 
stage  of  a  planned  research  project,  in  the  course  of  which  it  is  proposed  to  study 
these  rocks  and  their  faunas  over  the  whole  of  Britain. 

The  lower  Kimmeridgian  fauna  of  Britain  is  comparatively  well  known  and  it  was, 
therefore,  considered  that  the  best  way  to  carry  out  this  work  was  to  begin  at  the 
base  of  the  Upper  Kimmeridgian  (the  base  of  the  present  Gravesia  Zones)  and  work 
upwards  to  the  Portlandian.1  The  scope  of  the  first  part  of  the  work  embraces  the 
Gravesia,  "  Subplanites  "  and  basal  Pectinatites  Zones  of  the  type-section. 

1  The  Middle  Kimmeridgian  of  Arkell  (1956  :  21)  has,  as  a  result  of  the  conclusions  presented  herein, 
no  real  standing.  It  seems  most  appropriate  to  have,  therefore,  two  subdivisions  of  the  Kimmeridgian 
Stage,  the  line  between  them  being  drawn  at  the  top  of  the  Aulacostephanus  autissiodorensis  Zone. 


UPPER  KIMMERIDGE  CLAY  OF  DORSET  5 

Prior  to  1913  most  of  the  ammonites  from  the  British  Upper  Kimmeridgian  were 
known  under  the  name  of  Ammonites  biplex — a  name  which  embraced  practically  all 
Upper  Jurassic  perisphinctid  ammonites. 

In  1913  Salfeld  identified  some  of  the  British  Upper  Kimmeridgian  ammonites 
with  Pavlow's  genus  Virgatites.  Neaverson  (1925)  showed  that  these  ammonites 
were  not  related  to  the  Russian  Volgian  forms  as  Salfeld  had  supposed.  Neaverson's 
work,  valuable  though  it  was,  fell  far  short  of  monographing  the  whole  ammonite 
fauna  from  these  beds.  Neaverson  did  not  undertake  comprehensive  collecting 
from  the  type-section  in  Dorset,  which  surely  must  be  the  basis  for  zonal  subdivision. 
The  sequence  there  is  a  thick  one,  and  as  far  as  is  known,  complete.  In  contrast, 
the  sections  in  the  Oxford  district,  taken  by  Neaverson  as  the  basis  for  his  zonal 
scheme,  are  incomplete  and  very  much  attenuated.  As  a  result,  parts  of  his  zonal 
scheme  are  unsatisfactory. 

Spath  in  the  early  1930's  made  a  collection  from  the  Kimmeridge  section  (now 
housed  in  the  British  Museum  (Natural  History)),  and  although  this  was  never 
described,  references  are  made  to  it  in  several  of  his  papers  (e.g.  1935  :  73).  He 
identified  Neaverson's  genera  Allovirgatites  and  Virgatosphinctoides  with  his  genus 
Subplanites,  proposed  in  1925  for  a  group  of  ammonites  occurring  in  Franconia; 
he  later  identified  other  British  forms  with  the  Tithonian  genus  Lithacoceras;  and 
as  a  result,  correlations  became  established  between  Britain  and  Germany  based  on 
these  genera.  These  correlations  were  followed  among  others  by  Arkell  (1956), 
and  have  now  become  generally  accepted. 

The  collections  from  Dorset  now  to  be  described,  however,  establish  beyond  doubt 
that  the  British  Upper  Kimmeridgian  and  German  Tithonian  forms  are  not  identical, 
and  these  previous  correlations  are  thus  without  real  value. 

Many  colleagues  and  friends  have  been  of  invaluable  assistance  in  providing  helpful 
suggestions  and  criticisms.  I  am  particularly  indebted  in  this  respect  to  Professor 
D.  T.  Donovan,  Dr.  J.  H.  Callomon,  and  Dr.  A.  Zeiss. 

The  receipt  of  a  Research  Studentship  from  the  former  Department  of  Scientific 
and  Industrial  Research,  a  grant  from  the  British  Council  under  the  Younger  Research 
Workers  Interchange  Scheme,  and  financial  assistance  from  the  University  College 
of  Swansea  are  gratefully  acknowledged. 

II.     THE    KIMMERIDGE   SECTION 

The  Kimmeridge  Clay  is  the  oldest  formation  exposed  on  the  Isle  of  Purbeck.  It 
appears  as  a  long  strip,  approximately  six  miles  in  length  and  usually  less  than  a  mile 
wide,  in  the  core  of  the  Purbeck  Anticline.  To  the  north  the  steep  escarpment  of 
the  Portland  Stone  effectively  isolates  this  relatively  low-lying  land.  The  village 
of  Kimmeridge  is  situated  in  a  hollow  beneath  this  escarpment. 

The  outcrop  of  the  Kimmeridge  Clay  on  the  northern  limb  of  the  Purbeck  Anticline 
occurs  in  a  military  zone  extending  five  miles  westwards  from  Kimmeridge,  and  is 
for  this  reason  inaccessible.  Eastwards  from  Kimmeridge  Bay,  however,  the  section 
is  well  displayed  for  over  three  miles  on  the  southern  limb  of  the  anticline.  The  sea 
erodes  at  a  substantial  rate  the  relatively  soft  shales  which  comprise  the  bulk  of  the 
succession,  but  has  little  effect  on  the  occasional  cementstone  bands  which  stretch 


8  UPPER  KIMMERIDGE  CLAY  OF  DORSET 

Blake's 

Bed 
Number  //.         in. 

Pectinatites  (Arkellites)  hudlestoni  Zone 

20    pars       Shales,     mudstones,  hard "  dicey  "  bands      .  .  .51         8 

-23  Pectinatites  (Virgatosphinctoides)  encombensis,  P.  (V.) 

magnimasculus ,    P.    (Arkellites)    hudlestoni,    Lucina 

miniscula,    Protocardia    morinica,     Ostrea     bononiae, 

Discina  latissima,  Ichthyosaurus  sp. 

24  Basalt  Stone  Band       .......       3         10 

Inoceramus  expansus,  Ostrea  sp. 

25  "  Dicey"mudstones     .......     55  0 

Pectinatites    (Virgatosphinctoides)  donovani,  Inoceramus 

expansus,  Parallelodon  sp. 

26  pars        Shales       .........     13  1 

Pectinatites   (Arkellites)      hudlestoni,      P.      (Virgato- 
sphinctoides) reisiformis ,  P.  (V.)  reisiformis  densicos- 
tatus,  P.  (V.)  cf.  reisiformis,  P.  (V.)  abbreviatus,P.  (V.) 
grandis  acceleratus,  Ostrea  bononiae,  Ostrea  multiformis, 
Ostrea  sp.,  Protocardia  sp.,  Discina  latissima. 
26  pars        Cementstone      ........  4 

26  pars        Shales       .........       8         11 

Pectinatites    (Virgatosphinctoides)    reisiformis,    Ostrea 
bononiae,  Ostrea  sp. 
26  pars        Rope  Lake  Head  Stone  Band       .....       1  6 

Pectinatites  (Virgatosphinctoides)  wheatleyensis  Zone 

26  pars        Shales       .........     14  7 

Pectinatites  (Virgatosphinctoides)  reisiformis,  P.  (V.) 
wheatleyensis  delicatulus ,  P.  spp.  indet.,  Ostrea  sp. 

26  pars        The  Blackstone,  or  Kimmeridge  Oil  Shale    .  .  .        2         10 

Pectinatites  (Virgatosphinctoides)  wheatleyensis  delica- 
tulus, P.  spp.  indet,  Saccocoma  sp.,  Ostrea  sp.,  Discina 
latissima,  Lepidotus  sp. 

27  pars        Shales     .........       13  6 

Pectinatites  (Virgatosphinctoides)  wheatleyensis,  P.  (V.) 
wheatleyensis    delicatulus,    P.    (V.)    grandis,  P.    (V.) 
pseudoscruposus,  P.   (V.)  cf.  pseudoscruposus,  P.  (V.) 
laticostatus ,  P.  (V.)  woodwardi,  Saccocoma  sp.,  Ostrea 
bononiae,    Protocardia    morinica,    Lucina      miniscula, 
Discina  latissima,  Lingula  ovalis,  Cerithium  sp. 
27  pars        Siltstone  .........  1 

27  pars        Shales       .........       2  3 

Pectinatites  (Virgatosphinctoides)  laticostatus,  P.  (V.) 
woodwardi,  P.  (V.)  wheatleyensis,  P.  (V.)  grandis. 


UPPER    KIMMERIDGE    CLAY    OF   DORSET 


Blake's 

Bed 
Number 

27  pars 


27  pars 
28-30 

31 

32-33 
32 

33 

34 

35 


ft- 


36 


37- 
41  pars 


Siltstone  ......... 

Pectinatites   (Virgatosphinctoides)   laticostatus,   Discina 

latissima. 
Shales       ......... 

Pectinatites     {Virgatosphinctoides)     grandis,     P.     (V.) 

wheatleyensis  minor,  Ostrea  sp. 
"  Dicey  "  mudstones  and  shales  ..... 

Pectinatites   (Virgatosphinctoides)   laticostatus,  P.    (V.) 

smedmorensis,  P.  (V.)  clavelli,  Ostrea  bononiae,  Proto- 

cardia  morinica,  Lingula  ovalis. 
Grey  Ledge  Stone  Band  (Top  Ledge  of  Spath) 

Pectinatites  (Virgatosphinctoides)  scitulus  Zone 

Upper  Cattle  Ledge  Shales 

"  Dicey  "  mudstones  ...... 

Ostrea  sp. 

Shales       ......... 

Ophthalmosaurus  sp. 

Cattle  Ledge  Stone  Band    ...... 

Pectinatites  (Virgatosphinctoides)  cf.  major. 

Lower  Cattle  Ledge  Shales  ..... 

Pectinatites  (Virgatosphinctoides)  scitulus,  P.  (V.)  cf. 
scitulus,  P.  (V .)  decor osus,  P.  (V .)  major,  P.  (V.)  cf. 
major,  P.  (Arkellites)  cuddlensis,  P.  (A.)  damoni, 
Gravesia  cf.  gravesiana,  Ostrea  bononiae,  0.  multiformis, 
Protocardia  morinica,  Lucina  miniscula,  Modiola  autis- 
siodorensis,  Exogyra  virgula,  Lingula  ovalis,  Discina 
latissima,  Cerithium  sp.,  Pliosaurus  sp.,  Thrissops  sp. 

Yellow  Ledge  Stone  Band  ...... 

Pectinatites  (Arkellites)  cuddlensis,  P.  (Virgatosphinc- 
toides) elegans  corniger,  P.  (V.)  cf.  scitulus. 

Pectinatites  (Virgatosphinctoides)  elegans  Zone 

Hen  Cliff  Shales  ....... 

Pectinatites  (Virgatosphinctoides)  elegans,  P.  (V.)  elegans 
corniger,  P.  (V.)  major,  P.  (V.)  cf.  major,  P.  (Arkellites) 
cuddlensis,  P.  (A.)  primitivus,  P.  sp.  indet.,  Gravesia 
gigas,  G.  cf.  gravesiana,  Ostrea  bononiae,  Ostrea  multi- 
formis, Ostrea  sp.,  Exogyra  virgula,  Trigonia  pellati, 
Protocardia  morinica,  Lucina  miniscula,  Modiola 
autissiodorensis ,  Lingula  ovalis. 


in. 

I 


10 


45 


17 


5i 


69 


io  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Blake's 
Bed 
Number  ft.  in. 


41  pars        Cementstone 

41  pars        Shale 

Pectinatites  (Arkellites)  cf 

42  Cementstone 

Pectinatites  (Arkellites)  cf 

Total  thickness 


primitivus 
primitivus 


1  0 

1         11 

8 
.   464  6 


III.    the  non-ammonite  fauna 


The  non-ammonite  fauna,  although  often  abundant  in  terms  of  individuals,  is 
represented  by  few  species. 

VERTEBRATA 

Pisces.  Fish  remains  are  common  throughout  the  succession,  but  consist  mainly 
of  isolated  scales.  Fish  scales  are  exceedingly  abundant  in  the  Pectinatus  Zone. 
Identifiable  fish  remains  include  Thrissops  sp.,  and  Lepidotus  sp. 

Reptilia.  Vertebrae  and  occasionally  other  bones  occur  quite  commonly. 
The  anterior  part  of  a  skeleton  of  Ophthalmosaurus  was  found  12  ft.  above  the  Cattle 
Ledge  Stone  Band,  and  a  skull  of  Ichthyosaurus  from  12  ft.  below  the  White  Stone 
Band.  A  posterior  tooth  of  a  Pliosaur  was  found  5  ft.  above  the  Yellow  Ledge 
Stone  Band. 

BRACHIOPODA 

Discina  latissima  (Sow.)  occurs  throughout,  but  is  more  common  above  the  Black- 
stone  Band. 

Lingula  ovalis  Sow.  also  ranges  through  the  succession,  but  appears  to  reach  its 
maximum  just  above  the  Yellow  Ledge  Stone  Band. 

ECHINODERMATA 

One  species  of  crinoid  (Saccocoma  sp.)  occurs  as  isolated  pyritized  plates.  It 
appears  to  be  confined  to  the  Blackstone  and  the  ten  feet  of  shale  immediately  below. 

MOLLUSC  A 

Gastropoda:  Species  of  Cerithium  have  been  found  27  ft.  above  the  Yellow  Ledge 
Stone  Band,  and  3  ft.  below  the  Blackstone;  apart  from  these  occurrences  no  other 
gastropods  were  recorded. 

Bivalvia:  Bivalves  are  the  most  abundant  of  the  non-ammonite  fauna.  Exogyra 
virgula  (Defrance)  occurs  up  to  27  ft.  above  the  Yellow  Ledge  Stone  Band. 

Protocardia  morinica  (de  Loriol)  and  Lucina  miniscula  Blake  are  abundant  through- 
out. 


UPPER  KIMMERIDGE  CLAY  OF  DORSET  n 

Ostrea  bononiae  Sauvage  and  0.  multiformis  Koch  &  Dunker,  commonly  occur 
attached  to  the  undersurfaces  of  ammonite  shells. 

Other  bivalves  include  Trigonia  pellati  Mun-Chal.,  Parallelodon  sp.,  Inoceramus 
expansus  Blake,  Ostrea  solitaria  Sow.,  and  Modiola  autissiodorensis  (Cott.) 

Nautiloids  and  belemnites  are  absent  from  the  fauna. 

IV.     THE    AMMONITE    FAUNA 

(a)     Introduction 

The  state  of  preservation  of  the  ammonites  from  the  Kimmeridge  section  leaves 
much  to  be  desired  from  the  point  of  view  of  palaeontological  investigations. 

The  nodule  bed  in  the  Rotundum  Zone  of  the  Kimmeridge  Clay  is  well  known 
as  one  horizon  which  yields  reasonably  well  preserved  uncrushed  ammonites.  The 
author  has  found  uncrushed,  or  relatively  uncrushed,  specimens  at  two  other  hori- 
zons, both  of  them  in  the  part  of  the  section  described  herein.  One  large  isolated 
nodule  25  ft.  below  the  Yellow  Ledge  Stone  Band  yielded  a  few  pyritized  ammonite 
nuclei,  none  of  which  are  determinable  specifically.  The  other  horizon,  the  roof  bed 
of  the  Blackstone,  yields  ammonites  preserved  in  solid  pyrite,  but  in  which,  un- 
fortunately, the  septa  have  been  completely  pyritized  and  have  become  destroyed. 

All  the  other  ammonites  from  other  horizons  have  suffered  crushing  to  a  high 
degree.  The  ribbing,  however,  is  generally  well  preserved,  and  is  the  basis  for 
identification.     The  suture  has  almost  invariably  been  completely  obliterated. 

Considerable  problems  have  had  to  be  surmounted  both  in  the  collection  and  the 
preparation  of  these  ammonites.  Fossils  cannot  be  collected  from  the  cliffs,  owing 
to  the  fissile  nature  of  most  of  the  rock,  and  the  consequent  weathering  of  the  shale 
along  the  bedding.  The  abundant  pyrite  has  oxidized,  and  the  resulting  selenite 
crystals  cover  the  surface  of  the  bedding  planes. 

The  shale  reefs  exposed  at  the  base  of  the  beach  at  low  tide  are  the  only  places 
where  ammonites  can  be  satisfactorily  collected. 

At  some  horizons,  because  of  the  very  closely-spaced  joints  in  the  mudstone  bands, 
it  is  impossible  to  extract  the  ammonites.  In  this  case  a  plaster  cast  of  the  ammonite 
impression  is  made  in  the  field.  The  detail  reproduceable  with  thinly-mixed  plaster 
is  excellent,  and  the  casts  so  obtained  are,  for  most  purposes,  as  satisfactory  to 
work  with  as  the  actual  ammonites  from  other  horizons. 

The  ammonites  as  they  are  extracted  from  the  rock  form  most  unpromising-looking 
material  from  the  palaeontological  point  of  view,  and  careful  preparation  is  necessary 
before  sufficient  detail  is  visible  for  any  determinative  work. 

The  rib  interspaces  are  filled  with  hard  shale,  and  often  the  whole  ammonite  is 
encrusted  with  irregular  pyrite  aggregates.  The  lower  surface  of  the  ammonite 
(lower  surface  of  the  ammonite  as  it  lies  in  the  rock)  is  very  often  more  or  less 
encrusted  with  oysters  which  are  impossible  to  remove  successfully. 

The  crushing  of  the  ammonites  has  affected  most  of  the  original  measurements. 
The  diameter  has  been  increased  by  the  flattening  of  the  outer  whorls;  the  whorl 
height  is  similarly  affected,  whilst  the  whorl  thickness  is  reduced  to  about  one-eighth 
of  the  original  dimension.     The  diameter  of  the  umbilicus  is,  however,  relatively 


12  UPPER   KIMMERIDGE   CLAY   OF   DORSET 

unchanged.  The  crushing  of  the  outer  whorl  has  the  effect  of  making  the  point  of 
bifurcation  of  the  ribs  appear  much  lower  on  the  whorl  side  than  it  is  in  reality. 

For  purposes  of  identification  and  speciation,  therefore,  the  conventional  four 
measurements  are  not  given  (i.e.  diameter,  then  the  other  three  measurements  ex- 
pressed as  a  percentage  of  the  diameter).  Instead  the  diameter  and  the  umbilical 
diameter  only  are  given.  These  are  both  given  as  measurements  since,  as  mentioned 
above,  the  diameter  has  been  increased  by  the  crushing. 

Also  given,  where  possible,  are  the  number  of  primary  and  secondary  ribs  on  the 
outer  whorl,  and  the  number  of  ribs  at  various  diameters  (usually  at  5  mm.  intervals) 
within  the  umbilicus. 

In  most  cases  the  macroconch  of  a  species  is  designated  as  the  holotype.  Where 
this  is  not  possible  (i.e.  when  the  macroconch  of  a  species  has  not  been  found,  or 
when  no  well-preserved  macroconch  has  been  obtained),  the  microconch  is  designated 
as  the  holotype. 

In  cases  where  the  collection  of  specimens  can  only  be  carried  out  by  the  making 
of  plaster  casts  in  the  field,  the  macroconch  casts  have,  not  infrequently,  very  much 
obscured  inner  whorls.  The  reason  for  this  is  that  the  casts  are  made  of  the  under- 
surface  of  the  ammonites  as  they  lie  in  the  rock,  and  these  under-surfaces  are  often 
encrusted  with  oysters. 

(b)    Systematic  descriptions 

Phylum  Mollusca 

Class  CEPHALOPODA 

Sub-class  AMMONOIDEA 

Order  AMMONITIDA 

Superfamily  PERISPHINCTACEAE 

Family  PERISPHINCTIDAE 

Sub-family  AULACOSTEPHANINAE  Spath  1924 

Genus  GRAVESIA  Salfeld  1913 

Type  species  by  subsequent  designation  (Roman  1938):  Ammonites  gravesianus 
d'Orbigny  1850. 

Gravesia  gigas  (Zieten) 

(PI.  1,  fig.  1) 
1830     Ammonites  gigas  Zieten,  pi.  13,  fig.  1. 
1963     Gravesia  gigas  (Zieten)  ;   Hahn  :   97,  pi.  9,  pi.  10,  figs.  1,  2  (see  also  for  earlier  references). 

Material.     Two  specimens. 

Stratigraphical  range.     40-45  ft.  below  the  Yellow  Ledge  Stone  Band. 
Description.     These   two   specimens,   which   are   similar   to   one   another,   are 
crushed  flat.     Diameter  308-322  mm.     Diameter  of  umbilicus  104-114  mm.     The 


UPPER  KIMMERIDGE  CLAY  OF  DORSET 


13 


RIB    DIRECTION 


RURSIRADIATE 


RECTIRADIATE 


PRORSI  RADIATE 


RIB    TYPE 


SECONDARY 


PRIMARY 


INTERCALATORY 


BIFURCATE 


SIMPLE 


POLY  GYRATE 


POLYPLOKE 


VIRGATOTOME 


Fig.  2.     Rib  directions  and  rib  types  occurring  in  the  ammonites  of  the 
Upper  Kimmeridge  Clay. 


i4  UPPER    KIMMERIDGE   CLAY   OF   DORSET 

original  diameter  was  probably  around  240  mm.  There  is  a  very  close  resemblance 
to  the  neotype  (Hahn  1963,  pi.  9,  fig.  1),  the  only  discernible  difference  being  that 
the  Dorset  specimens  still  have  quite  prominent  secondary  ribs  on  the  venter  at  the 
aperture. 

Remarks.  This  species  is  extremely  rare  in  Dorset,  two  other  specimens  exist  in 
the  collections  of  Spath  in  the  British  Museum  and  are  also  from  the  same  horizon. 
Salfeld  recorded  "  numerous  examples  "  of  G.  gravesiana  from  about  this  horizon 
(Arkell  1933  :  440).     This  species  appears,  however,  to  occur  higher  in  the  succession. 


Gravesia  cf.  gravesiana  (d'Orbigny) 

(PI.  1,  fig.  2) 

1850     Ammonites  gravesianus  d'Orbigny  :    559,  pi.  219,  figs.  1,  2. 

1963     Gravesia  gravesiana  (d'Orbigny)  ;    Hahn  :    99,  pi.  10,  figs.  3,  4  ;    pi.  12,  figs.  3,  4  ;    pi.  13, 
fig.  2.     (See  also  for  earlier  references.) 

Material.     Two  specimens. 

Strati  graphical  range.  From  8  ft.  below  to  6  ft.  above  the  Yellow  Ledge 
Stone  Band. 

Description.  Both  the  specimens  are  whorl  fragments.  The  larger  (PI.  1,  fig.  2) 
shows  the  ribbing  very  well  and  is  very  close  to  G.  gravesiana.  The  total  estimated 
crushed  diameter  of  this  specimen  would  be  about  160  mm. 

The  other  specimen  shows  only  three  primary  ribs,  with  internal  moulds  of  the 
secondary  ribs,  but  is  similar  in  rib  style  to  the  first  specimen. 

Remarks.  Salfeld  recorded  Gravesia  irius  at  about  this  horizon,  a  species  which 
the  author  was  not  able  to  find. 

The  occurrence  of  Gravesia  above  the  Yellow  Ledge  Stone  Band  has  not  hitherto 
been  reported,  but  as  this  genus  is  so  rare  in  Dorset  it  is  by  no  means  certain  that 
the  newly  established  range  of  60  feet  in  Dorset  is  the  maximum  range  of  this  genus. 


Subfamily  VIRGATOSPHINCTINAE  Spath  1923 

Genus  PECTIN ATITES  Buckman  1922 
Type  species  Ammonites  pectinatus  Phillips  1,871. 

Diagnosis.  Dimorphic.  Microconchs  generally  65-110  mm.  in  diameter, 
occasionally  larger.  Inner  whorls  with  sharp  biplicate  ribbing,  becoming  a  little 
coarser  on  body-chamber  with  occasional  simple  and  trifurcate  ribs.  Peristome 
with  ventral  horn.  Body-chamber  generally  half  a  whorl  long.  Macroconchs 
generally  140-200  mm.  occasionally  larger,  very  rarely  smaller.  Inner  whorls 
with  sharp  biplicate  ribbing.  Outer  whorl  very  variable,  usually  with  strong  primary 
ribs  and  two  to  five  secondary  ribs  to  each  primary  rib.  Peristome  simple.  Body- 
chamber  generally  half  a  whorl  long.     Constrictions  present  in  some  species. 

Upper  Kimmeridgian.     (Elegans  to  Pectinatus  Zones). 


UPPER    KIMMERIDGE   CLAY    OF    DORSET  15 

Sexual  Dimorphism 

Detailed  collecting  from  the  Upper  Kimmeridge  Clay  at  the  type-section  at 
Kimmeridge,  Dorset,  has  revealed  many  ammonites  referable  to  species  of  the 
genus  Pectinatites  Buckman.  Random  sampling  has  established  that  the  vast 
majority  of  these  fall  into  one  of  two  size  groups.  Formerly  it  has  been  considered 
by  those  who  had  collected  ammonites  from  this  section  (e.g.  Arkell  &  Spath),  that 
the  smaller  specimens  were  merely  young  forms  of  the  larger.  However,  these 
previous  collections  consist  mostly  of  individuals  from  which  the  peristome  had  been 
broken  during  extraction  from  the  rock,  and  examination  of  recent  collections, 
consisting  mainly  of  individuals  with  peristome  intact,  suggests  a  new  interpretation 
of  the  size  grouping.  This  interpretation  arises  from  the  fact  that  at  every  horizon 
from  which  collections  were  made,  only  two  size  groups  are  found.  The  smaller 
size  falls  into  the  65-110  mm.  diameter  range,  and  the  larger  into  the  140-200  mm. 
diameter  range.  If  the  ammonites  of  the  smaller  size  group  were  the  young  of  the 
larger  ones  it  would  be  remarkable  not  to  find,  at  some  horizon,  ammonites  falling 
into  the  size  range  110-140  mm.  diameter.  Evidence  to  show  that  the  two  groups 
are  quite  distinct  follows  below. 

Callomon  (1963  :  25)  has  summarized  the  criteria  by  which  an  ammonite  may  be 
judged  to  be  mature.     These  are: 

(a)  Uncoiling  of  the  umbilical  seam. 

(b)  Modification  of  sculpture  near  the  peristome;   usually  a  coarsening 

and  re-  or  degeneration  of  ribbing,  but  often  with  terminal  constric- 
tions, ventral  collars,  flares,  horns,  rostra,  lateral  lappets  etc. 

(c)  Approximation  and  degeneration  of  the  last  few  septal  sutures. 

As  the  ammonites  are  badly  crushed,  all  traces  of  the  septa  have  been  destroyed. 
If,  however,  the  first  two  of  these  criteria  are  applied,  it  is  found  that  both  size  groups 
mentioned  above  consist  of  mature  individuals. 

Species  of  these  perisphinctid  ammonites  are  best  distinguished  from  one  another 
by  the  density  of  the  ribbing.  If  the  numbers  of  ribs  at  given  diameters  are  plotted 
against  diameters  on  a  graph,  a  curve  is  produced  which  is  distinctive  for  any  given 
species.  The  two  size  groups  under  consideration  here  give  generally  similar,  but 
not  identical  curves,  so  that  it  is  possible  to  distinguish,  by  means  of  rib  curves,  the 
larger  from  the  smaller  type,  even  with  incomplete  material.  These  two  groups 
have  been  referred  to  as  microconchs  and  macroconchs  by  Callomon  (1957  :  62),  a 
terminology  which  has  become  generally  accepted. 

The  microconchs,  in  this  case,  are  generally  small  forms  with  a  diameter  of  65-110 
mm.,  but  at  one  horizon  they  range  up  to  185  mm.  in  size.  In  all  cases  the  ribbing 
is  of  normal  perisphinctid  biplicate  style  and,  apart  from  slight  coarsening  towards 
the  aperture  and  occasional  development  of  polygyrate  ribs,  shows  little  modification. 
The  aperture  bears  a  horn-like  process  projecting  from  the  venter.  The  umbilical 
seam  gradually  uncoils  over  the  last  half  whorl,  so  that  at  the  aperture  many  forms 
are  completely  evolute.  The  apertural  margin  is  sometimes  devoid  of  ribbing  and 
shows  a  smooth  zone,  ornamented  little,  save  for  growth  lines,  and  the  presence  of 
the  ventral  horn. 


16  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

The  macroconchs  are  usually  140-200  mm.  in  diameter,  but  occasionally  are  larger, 
or  very  rarely  smaller,  and  are  characterized  by  a  smooth  sinuous  peristome  margin. 
The  ribbing  on  the  inner  whorls  is  of  a  simple  biplicate  style,  but  the  point  of  bifurca- 
tion usually  occurs  slightly  higher  on  the  whorl  side  than  it  does  on  the  microconchs. 
The  body-chamber  develops  irregular  ribbing  and,  particularly  in  the  forms  from 
higher  horizons,  has  a  tendency  to  fasciculation  or  virgatotomy.  Uncoiling  of  the 
umbilical  seam  occurs  only  over  the  last  half  whorl. 

The  two  forms  are  found  in  association  throughout  the  succession,  although  the 
ratio  of  microconchs  to  macroconchs  varies.  This  ratio  is  usually  within  the  limits 
2  :  1-1  :  2.  Where  only  a  few  specimens  have  been  obtained  from  one  horizon 
this  ratio  is  not  treated  as  significant. 

That  these  two  groups  of  ammonites  are  very  closely  related  is  strongly  suggested 
by  their  co-existence  at  each  fossiliferous  horizon,  their  identical  stratigraphical 
range,  and  their  similar  rib  curves.  However,  four  specimens  from  one  horizon 
(13  feet  above  the  Rope  Lake  Head  Stone  Band)  from  which  have  been  collected 
32  microconchs  and  34  macroconchs,  show  conclusively  the  relationship.  Three  of 
these  specimens  appear  to  be  normal  macroconchs,  but  have  on  their  inner  whorls 
structures  resembling  those  of  the  horn  of  the  microconch.  However,  this  structure 
differs  from  the  true  microconch  horn;  it  has  negligible  ventral  projection,  it  is 
developed  from  a  single  rib,  and  it  projects  laterally. 

The  fourth  of  these  specimens  is  unique  in  that  it  is  intermediate  in  size  between 
the  two  groups  (117  mm.  diameter),  has  the  typical  microconch  horn  developed, 
but  shows  the  beginnings  of  the  macroconch  type  of  ribbing  associated  with  four 
further  horns.  The  rib  density  of  the  first  three  of  the  ammonites  shows  them  to 
have  affinity  with  the  macroconch  group.  The  fourth  specimen  has  a  rib  density 
intermediate  between  that  of  a  microconch  and  a  macroconch. 

The  undersides  of  the  ammonites,  as  they  lie  in  the  rock,  are  quite  often  encrusted 
with  oysters,  although  the  upper  surface  is  generally  free  of  them.  It  would,  there- 
fore, seem  that  the  oysters  attached  themselves  to  the  ammonite  conch  after  the 
death  of  the  latter,  otherwise  the  oysters  would  presumably  be  equally  common  on 
both  surfaces.  They  apparently  grew  in  the  shelter  provided  by  the  umbilical  space 
beneath  the  ammonite,  and  flourished  there  until  continued  sedimentation  eventually 
killed  them.  Medcof  (1955)  has  shown  that  modern  oyster  larvae  prefer  to  settle  on 
under-surfaces.  In  this  case  the  ammonite  shells  would  provide  the  only  such 
surfaces  available  on  the  sea  bed. 

Judging  by  the  size  of  these  oysters,  a  considerable  time  must  have  elapsed  before 
they  were  killed  by  the  continued  influx  of  sediment,  so  that  we  may  reasonably 
conclude  that  sedimentation  was  not  rapid.  This  is  also  supported  by  lithological 
evidence.  The  rocks  are  a  fairly  uniform  argillaceous  series — grey  and  black  shales 
and  clays  with  occasional  cementstone  bands — and,  apart  from  lamination,  are 
devoid  of  sedimentary  structures. 

Save  for  the  very  occasional  juvenile  forms  and  occasional  gerontic  forms,  all  the 
ammonites  fall  into  one  of  the  two  size  groups  mentioned  previously,  and  bearing  in 
mind  the  evidence  of  slow  deposition,  it  is  likely  that  the  ammonite  faunas  of  the 
Upper  Kimmeridge  Clay  in  Dorset  represent  a  death  assemblage. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  17 

Taking  into  account  the  evidence  of  maturity  of  the  ammonites,  and  the  fact 
that  they  represent  death  assemblages,  it  would  appear  evident  that  the  difference 
in  size  of  the  two  groups  is  of  a  fundamental  nature.  It  seems  most  unlikely  that 
current  sorting  of  the  shells,  or  sudden  extermination  of  whole  populations  occurred. 
The  most  obvious  interpretation  of  this  size  distribution  is  that  these  ammonites 
exhibit  dimorphism.  Dimorphism  of  ammonite  shells  is  probably  an  expression  of 
some  difference  which  was  present  in  the  soft  parts  also.  The  most  obvious  differ- 
ence between  dimorphs  would  appear  to  be  a  sexual  one,  and  there  is  some  evidence 
to  suggest  that  the  microconch  and  macroconch  may  represent  the  two  sexes. 

Examination  of  the  microconchs  yields  several  important  facts  relating  to  the  horn. 
It  is  never  developed  until  a  diameter  of  at  least  60  mm.  (generally  more)  is  attained. 
In  other  words,  the  horn  is  not  developed  until  a  certain  stage  of  growth  is  reached. 
At  various  growth  stages  beyond  this  diameter  further  horns  may  be  developed,  but 
the  presence  of  a  former  horn  or  horns  is  always  detectable.  Sometimes  the  earlier 
horn  is  retained,  and  in  other  cases  the  earlier  horns  appear  to  have  been  shed,  and  to 
have  left  behind  a  characteristic  scar  on  the  venter. 

Apart  from  the  four  macroconch  specimens  mentioned  earlier,  none  shows  any 
trace  on  the  earlier  whorls  of  any  type  of  horn  or  ventral  scar. 

In  section,  the  microconch  horn  is  U-shaped,  opening  forwards.  This  suggests 
that  it  housed  some  part  of  the  soft  parts  of  the  animal,  and,  since  the  horn  is  confined 
to  the  microconch,  it  is  reasonable  to  assume  that  its  function  may  have  been  sexual. 
This  would  explain  its  confinement  to  the  microconch,  and  its  occurrence  only  in 
nearly  full-grown  specimens. 

The  four  specimens  showing  characteristics  of  each  group  can  then  be  explained  as 
various  degrees  of  intersexual  specimens.  Three  of  them  are  barely  distinguishable 
from  true  macroconchs,  but  the  fourth  appears  to  be  a  true  intersex. 

As  mentioned  above,  the  macroconchs  and  microconchs  differ  somewhat  in  the 
density  of  their  ribbing.  At  15  mm.  diameter  (the  smallest  diameter  at  which  it  is 
practicable  to  count  the  ribs  accurately)  the  macroconchs  are  nearly  always  finer 
ribbed  than  the  corresponding  microconchs.  The  comparative  density  of  the  ribbing 
of  the  two  forms,  at  greater  diameters,  is  seen  to  vary  with  the  species  concerned. 
Presumably,  both  microconch  and  macroconch  reached  maturity  at  the  same  age, 
so  that  the  rate  of  growth  of  the  macroconch  must  have  been  greater  than  that  of 
the  microconch.  This  would  appear  to  explain  these  discrepancies,  since  growth 
rate  in  each  species  must,  to  perhaps  a  small  and  varying  extent,  have  had  an  effect 
on  the  density  of  the  ribbing. 

Our  knowledge  of  the  soft  parts  of  ammonites  is  almost  entirely  based  on  analogy 
with  modern  cephalopods,  particularly  Nautilus.  In  most  living  cephalopods  the 
male  of  the  species  is  smaller  than  the  female.  In  Nautilus,  however,  the  male  is 
slightly  broader-shelled  than  the  female,  the  extra  breadth  of  the  shell  being  utilised 
to  incorporate  the  male  copulatory  organs,  the  diameter  of  the  two  shells  is  approxi- 
mately equal.  It  appears,  therefore,  that  in  any  case  of  marked  dimorphism  in 
modern  cephalopods  the  male  is  the  smaller  sex,  and  it  therefore  appears  likely  that  in 
Pectinatites  the  microconch  represents  the  male  of  the  species.  If  this  were  so,  the 
horn  may  have  assisted  in  copulation.     If  the  spadix  (the  copulatory  organ  of  the 

GEOL.  15,  I.  2 


18  UPPER   KIMMERIDGE   CLAY    OF   DORSET 

male  cephalopods)  were  housed  within  the  horn,  by  insertion  of  the  horn  within  the 
venter  of  the  female  shell,  fertilization  of  the  ova  would  be  more  readily  assured. 

Dimorphism  has  been  reported  in  other  groups  of  ammonites  by  various  authors 
(e.g.  Callomon  1963,  Makowski  1962,  Westermann  1964).  In  many  of  the  reported 
instances,  the  microconch  aperture  bears  a  pair  of  lappets  developed  laterally. 
Lappets  are  not  present  in  any  ammonites  found  hitherto  from  the  Upper  Kim- 
meridge  Clay,  while  the  horn  of  the  Kimmeridgian  microconchs  is  apparently  unique 
to  this  group  of  ammonites.  These  horned  Kimmeridgian  forms  are  known  outside 
Britain  from  Northern  France  and  Greenland,  and,  as  they  appear  to  have  evolved 
rapidly,  promise  much  in  precise  correlations  within  this  Upper  Jurassic  faunal 
province. 

Interpretation  of  PECTIN  AT  ITES 

The  genus  Pectinatites  was  originally  proposed  for  a  few  closely  related  species 
from  the  Pectinatus  Zone  of  Oxfordshire.  The  Pectinatus  Zone  there  is  to  be 
correlated  with  the  rocks  between  the  White  Stone  Band,  and  the  base  of  the  Pavlovia 
rotunda  Zone  in  the  Dorset  succession. 

As  early  as  1896  Hudleston  (1896  :  322)  had  remarked  on  the  similarity  between 
Ammonites  pectinatus  and  the  pyritized  ammonites  which  occur  at  the  top  of  the 
Kimmeridge  Oil  Shale  or  Blackstone,  about  150  ft.  below  the  White  Stone  Band  in 
the  Kimmeridge  section. 

Buckman,  in  June  1925,  assigned  one  species  of  these  pyritized  ammonites  from 
the  Blackstone  to  a  new  genus,  Pectini for  mites  which  he  placed  six  hemerae  earlier 
than  his  Pectinatus  hemera,  but  Neaverson  (Dec.  1925)  did  not  accept  Buckman's 
findings,  and  placed  these  Blackstone  ammonites  in  the  genus  Pectinatites.  Spath 
(1936  :  18)  in  turn  placed  them  in  his  genus  Subplanites  proposed  in  1925. 

Herein,  I  place  these  ammonites  in  the  genus  Pectinatites,  in  which  I  recognize 
three  subgenera:  Pectinatites  (sensu  stricto);  Virgatosphinctoides ;  and  Arkellites 
subgen.  nov. 

Many  generic  attributions  have  been  given  to  species  of  Pectinatites  in  the  past. 
These  generic  names,  for  the  most  part,  belong  to  quite  distinct  genera  many  of 
which  do  not  occur  in  Britain,  while  others  of  them  are  either  junior  sjmonyms,  or  in 
some  cases  subgenera  of  Pectinatites.  As  much  confusion,  and  many  unreliable 
correlations  have  been  made  on  the  basis  of  misidentification  of  species  of  Pectinatites 
with  other  genera,  there  follows  a  discussion  of  these  genera  and  their  relationship, 
if  any,  to  Pectinatites. 

VIRGATITES  Pavlow  1892 

Type  species.  Ammonites  virgatus  von  Buch  1832.  (Subsequently  designated 
Douville  19x0.) 

This  genus  was  recorded  from  the  horizon  of  the  Kimmeridge  oil-shale  by  Salfeld 
(1913),  and  a  zone  of  V.  miatschkovensis  introduced  by  him  for  the  beds  between  the 
Gravesia  zones  and  his  zone  oiPerisphinctes  pallasianus  (=  modern  Pectinatus  Zone). 
The  genus  is  characterized  by  virgatotome  ribbing  on  the  inner  whorls,  sometimes 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  19 

reverting  to  simple  or  bifurcate  ribbing  on  the  body-chamber.  All  the  Dorset 
ammonites  from  this  part  of  the  Kimmeridge  Clay  have  normal  perisphinctid 
bifurcate  ribs  on  their  inner  whorls,  and  it  is  only  on  the  outer  whorl  that  virgatotome 
ribbing  may  develop.  There  is  now  no  doubt  that  Virgatites  is  much  younger  in 
age  than  these  Kimmeridge  forms,  and  appears  to  be  restricted  to  the  Volgian  faunal 
province  of  eastern  Europe. 


PSEUDOVIRGATITES  Vetters  1905 

Type  species.     Ammonites  scruposus  Oppel  in  Zittel  1868. 

Lamplugh,  Kitchin  &  Pringle  (1923  :  222)  recorded  the  occurrence  of  the  genus 
Pseudovirgatites  from  Dorset,  and  introduced  a  zone  of  Pseudovirgatites  to  include  the 
horizon  of  the  Blackstone  in  Dorset.  This  genus  is  often  homeomorphic  with  the 
genus  Pectinatites.  The  type  species  from  the  Lower  Tithonian  has  similar  rib-style 
on  its  outer  whorl  to  that  of  some  large  species  of  Pectinatites  (e.g.  P.  (Virgatosphinc- 
toides)  pseudoscrnposus  (Spath)).  Other  species  of  Pseudovirgatites ,  such  as  some 
of  those  recently  figured  by  Donze  &  Enay  (1961)  and  Michailov  (1964),  are  remark- 
ably homeomorphic  with  some  species  of  Pectinatites  (e.g.  P.  (P.)  inconsuetus  sp.  nov. 
See  p.  138,  PI.  30).  The  microconch  of  Pseudovirgatites  does  not,  however,  possess  a 
ventral  peristomal  horn  as  does  Pectinatites,  and  all  records  of  Pseudovirgatites  from 
Britain  would  appear  to  refer  to  homeomorphic  forms  of  Pectinatites. 


LITHACOCERAS  Hyatt  1900 

Type  species.     Ammonites  ulmensis  Oppel  1863. 

The  inner  whorls  of  Lithacoceras  generally  bear  fine  bifurcate  ribs  which  modify 
on  the  outer  whorl  of  the  macroconch  to  produce  in  the  type-species  widely-spaced 
blunt  primary  ribs,  each  giving  rise  to  up  to  eight  secondaries.  Some  species  of  the 
genus  reach  a  very  large  size.  There  is  often  a  considerable  degree  of  homeomorphy 
between  species  of  this  genus  and  species  of  Pectinatites.  Apart  from  peristomal 
differences,  the  microconchs  of  the  two  genera  can  be  very  similar,  and  the  middle 
whorls  of  a  macroconch  of  Lithacoceras  sometimes  very  closely  approach  the  ornament 
of  the  macroconch  of  Pectinatites.  This  homeomorphy  has  misled  many  workers  in 
the  past.  In  particular  the  species  of  Pectinatites  from  the  Hen  Cliff  Shales  have 
been  identified  in  the  past  as  Lithacoceras.     (e.g.  Arkell  1956  :  21). 


PECTINATITES  Buckman  1922 

Type  species.     Ammonites  pectinatus  Phillips  1871. 

The  name  Pectinatites  is  the  most  senior  available  name  for  the  British  Upper 
Kimmeridgian  ammonites  described  herein.  It  is  distinguished  from  all  other  genera 
which  are  to  varying  degrees  homeomorphic  with  it,  by  its  type  of  dimorphism.  As 
these  ammonites  with  horned  microconchs  form  a  closely  related  natural  group,  it  is 
here  proposed  to  include  all  such  dimorphic  forms  in  this  genus. 


20  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

WHEATLEYITES  Buckman  1923 

Type  species.     Wheatleyitcs  tricostulatus  Buckman  1923. 

This  genus  is  characterized  by  finely-ribbed  inner  whorls,  which  modify  to  produce 
an  outer  whorl  with  coarse  widely-spaced  prima^  ribs;  the  secondary  ribs  gradually 
fade  on  the  body-chamber.  Some  forms  of  Wheatleyites  are  homeomorphic  with 
species  of  the  Tithonian  genus  Pseudovirgatites .  Wheatleyites  is  here  regarded  as  a 
junior  synonym  of  Pectinatites,  it  being  a  name  applied  by  Buckman  to  macroconchs 
of  Pectinatites  having  this  particular  type  of  modification  of  the  ribs  on  the  outer 
whorls. 

SUBPLANITES  Spath  1925  (January) 

Types  species.     Virgatosphinctes  reisi  Schneid  1914. 

To  this  genus  belong  a  complex  of  forms  occurring  in  the  Tithonian  rocks  of  Europe. 
Characteristically  their  inner  whorls  bear  fine  bifurcate  ribs,  which  are  modified  on 
the  body-chamber  in  a  fashion  similar  to  that  which  obtains  in  many  species  of 
Pectinatites.  It  was  for  this  reason  that  many  of  the  British  species  of  Pectinatites 
were  long  considered  to  be  species  of  Subplanites.  Virgatosphinctoid.es  Neaverson 
(here  treated  as  a  subgenus  of  Pectinatites),  was  considered  a  junior  synonym  of 
Subplanites.  This  undetected  homeomorphy  led  to  the  establishment  of  a  number  of 
unreliable  correlations  between  Britain  and  Southern  Europe. 

The  microconchs  of  Subplanites  bear  lappets,  and  are  for  this  reason  easily  dis- 
tinguishable from  the  microconchs  of  Pectinatites  when  material  with  intact  peristomes 
is  available,  but  in  the  absence  of  specimens  with  peristomes  it  is  virtually  impossible 
to  distinguish  the  two  genera. 

All  the  British  forms  appear  to  belong  to  Pectinatites,  but  in  the  case  of  such  faunas 
as  those  from  Russia,  recently  described  by  Michailov  (1964),  it  is  not  possible  to 
determine  the  genus  of  ammonites  present  owing  to  the  incomplete  nature  of  the 
material. 

PECTINIFORMITES  Buckman  1925  (June) 

Type  species  (by  monotypy).     Pectiniformites  bivius  Buckman  1925. 

The  holotype  which  is  in  the  Dorset  County  Museum,  Dorchester,  is  a  pyritic 
cast  from  the  Blackstone.  The  pyrite  of  the  outer  whorl  has  in  places  reached  an  ad- 
vanced state  of  decomposition,  and  the  specimen  is  now  of  little  value.  There  is  also 
in  this  museum,  however,  a  cast  of  the  holotype  made  in  1925,  which  appears  to 
correspond  very  closely  in  dimensions  to  the  holotype,  and  which  is  a  better  specimen 
than  the  holotype  in  its  present  condition. 

Buckman  marked  on  his  plate  of  the  holotype  (1925,  pi.  568)  the  position  of  the 
last  visible  suture,  which  is  just  one  whorl  back  from  the  supposed  peristome. 
However,  the  type  of  preservation  in  the  Blackstone  (solid  pyrite)  generally  destroys 
all  trace  of  the  septa,  and  in  the  author's  opinion,  the  septum  marked  by  Buckman 
was  the  last  visible,  but  not  the  last  occurring  septum.  This  view  is  supported  by 
the  occurrence  of  better-preserved  ammonites  having  affinities  with  this  species 


UPPER    KIMMERIDGE   CLAY    OF    DORSET  21 

and  with  a  short  body-chamber.  (E.g.  Pectinatites  pectinatus  (Phillips)  Buckman 
1922,  pi.  354B,  which  shows  five-eighths  of  a  whorl  of  body-chamber;  and  the  length 
of  the  body-chamber  estimated  from  differences  in  the  degree  of  crushing  of  the 
Dorset  material,  which  suggests  a  body-chamber  length  of  between  three  and  five- 
eighths  of  a  whorl.  No  specimen  of  Pectinatites  is  known  to  the  author  with  a  body- 
chamber  as  much  as  one  whorl  in  length). 

The  ammonites  from  the  Blackstone  are  largely  uncrushed,  but  there  is  another, 
more  important,  difference  between  the  ammonites  from  this  horizon  and  other 
horizons  in  the  Dorset  succession.  In  the  Blackstone,  ammonites  of  all  growth 
stages  are  preserved,  from  very  small  nuclei  to  specimens  over  150  mm.  in  diameter. 
This  contrasts  with  other  horizons  where  mature  individuals  make  up  by  far  the 
greater  part  of  the  ammonite  fauna.  Furthermore,  no  ammonites  have  been 
collected  (or  preserved?)  in  the  Blackstone  with  intact  peristomes.  This  means  that 
it  is  not  possible,  in  the  case  of  the  smaller  specimens,  to  distinguish  macroconchs 
from  microconchs,  and  thus  that  the  interpretation  of  Pectininf or  mites  is  open  to 
doubt. 

Further,  the  Blackstone  has  hitherto  failed  to  yield  macroconchs  with  well- 
preserved  inner  worls,  so  that  the  holotype  of  Pectini for  mites  bivius  cannot  be  com- 
pared to  any  known  macroconch  specimen.  It  is,  therefore,  not  possible  to  determine 
to  which  subgenus  of  Pectinatites  this  species  belongs.  The  rib  density  is  such  that 
affinity  with  Arkellites  subgen.  nov.  is  unlikely  (approximately  55  ribs  at  30  mm. 
diameter).  It  may  possibly  therefore  be  consubgeneric  with  Virgatosphinctoides, 
but  there  appears  to  be  no  over-riding  reason  why  it  should  not  equally  be  placed  in 
Pectinatites,  sensu  stricto .  This  was  also  the  view  of  Neaverson  (1925  :  15)"  Buckman 
has  recently  instituted  a  new  genus  Pectiniformites  for  ammonites  of  the  pectinatus- 
type  from  this  facies  (the  oil  shales  of  Kimmeridge).  There  seems  to  be  no  justifica- 
tion for  this,  and  Pectiniformites  must  be  regarded  as  synonymous  with  Pectinatites  " . 

Pectiniformites  would  thus  become  a  junior  synonym  of  Pectinatites,  and  is  so  treated 
here. 

KERATINITES  Buckman  1925  (October) 

Type  species.     Keratinites  keratophorus  Buckman  1925. 

This  genus  was  introduced  by  Buckman  for  ammonites  from  the  Pectinatus  Zone 
having  a  peristome  bearing  a  ventral  horn.  These  forms  are  the  microconchs 
of  Pectinatites,  and  the  name  Keratinites  is  here  regarded  as  a  junior  synonym  of 
Pectinatites. 

VIRGATOSPHINCTOIDES  Neaverson  1925  (December)  :  11 

Type  species.     Virgatosphinctoides  wheatleyensis  Neaverson  1925. 

This  genus  is  characterized  by  finely  ribbed  inner  whorls  which  are  modified  on 
the  body-chamber  of  the  macroconch,  often  producing  polygyrate,  polyploke,  or 
virgatotome  ribbing.  The  genus  was  regarded  by  Spath  as  synonymous  with,  or 
at  the  most  subgenerically  different  from,  his  genus  Subplanites  proposed  a  few  months 
earlier  (see  above).     Systematic  collections  from  Dorset  have  now  established  that 


22  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Virgatosphinctoides  is  dimorphic.  The  microconchs  bear  a  ventral  horn  and  are 
never  seen  to  have  lappets.  Virgatosphinctoides  is  thus  easily  distinguished  from 
Subplanites  when  material  with  intact  peristomes  is  available.  The  presence  of  a 
horned  peristome,  however,  shows  that  Virgatosphinctoides  is  closely  related  to 
Pectinatites.  The  microconchs  of  the  two  are  sometimes  indistinguishable,  and  only 
the  characters  of  the  macroconchs  can  usefully  separate  the  two  forms.  For  this 
reason  Virgatosphinctoides  is  here  treated  as  a  subgenus  of  Pectinatites. 

ALLOVIRGATITES  Neaverson  1925  (December)  :  29 

Type  species.     Allovirgatites  woodwardi  Neaverson  1925. 

Neaverson's  basis  for  distinction  between  Virgatosphinctoides  and  Allovirgatites 
appears  to  have  been  based  almost  entirely  on  differences  in  the  septal  suture  of 
species  of  the  two  genera.  However,  the  rib-style  and  its  development  is  very  similar 
in  these  two  forms,  and  there  appears  to  be  little  justification  for  drawing  distinction 
between  them.  Neaverson  admitted  similarity  between  the  suture  lines  of  these 
two  genera  in  the  young  stages,  and  there  would  seem  little  doubt  that  differences 
between  his  described  forms  are  no  more  than  specific  differences.  Allovirgatites 
is  therefore  here  regarded  as  a  junior  synonym  of  Virgatosphinctoides. 

SUBDICHOTOMOCERAS  Spath  1925  (January) 

Type  species.     Subdichotomoceras  lamplughi  Spath  1925. 

This  genus  is  characterized  by  sharply  biplicate  ribbing  throughout  development, 
together  with  deep  constrictions  which  are  bordered  by  simple  ribs.  The  aperture 
is  without  lappets.  The  holotype  came  from  the  Eudoxus  Zone  of  Yorkshire,  and 
the  genus  does  not  appear  to  be  represented  in  higher  Kimmeridgian  deposits  in 
Dorset. 

SPHINCTOCERAS  Neaverson  1925  (December) 

Type  species.     Sphinctoceras  crassum  Neaverson  1925. 

Two  species  of  Sphinctoceras  were  described  by  Neaverson  from  the  Wheatleyensis 
Zone  of  Oxfordshire.  They  are  massive  inflated  forms  with  coarse  strong  biplicate 
ribs.  There  seems  little  doubt  that  Sphinctoceras  is  closely  related  to  Subdichoto- 
moceras, the  former  being  almost  certainly  the  macroconch  of  the  latter.  No 
specimens  of  Sphinctoceras  have  hitherto  been  found  in  Dorset,  but  the  genus  is 
mentioned  here  because  the  conservative  "  biplex  "  stock  to  which  it  belongs  gave 
rise  to  the  pavlovids  in  the  Pectinatus  Zone.  The  sharp  biplicate  ribbing,  the  very 
high  point  of  bifurcation  of  the  ribs,  and  the  absence  of  polygyrate  ribs  and  any 
marked  apertural  modification  make  identification  of  these  forms  with  more  coarsely- 
ribbed  species  of  Pectinatites  unlikely. 

The  origins  of  PECTINATITES 

The  origins  of  Pectinatites  are  rather  obscure,  but  there  is  one  feature  of  the  ribbing 
which  must  be  considered  of  great  value  in  deducing  the  origin  of  the  genus.     This 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  23 

is  the  presence  of  the  polygyrate,  and  more  rarely  the  polyploke  rib  type  (Geyer 
1961,  text-fig.  1).  This  type  of  ribbing,  which  is  first  found  in  some  Upper  Oxfordian 
ammonites,  is  the  first  new  character  in  perisphinctid  ornamentation  to  appear  since 
the  Bajocian,  and  it  therefore  appears  very  probable  that  all  ammonites  which  have 
this  rib  style  are  related. 

In  the  Lower  Kimmeridgian  genus  Ataxioceras,  the  development  of  polygyrate 
ribbing  reaches  its  extreme.  Ataxioceras  is  also  often  ribbed  in  a  most  irregular 
fashion,  a  character  which  is  evident  in  many  species  of  Pectinatites.  A  further 
character  of  Ataxioceras  is  of  importance  too  in  tracing  the  origin  of  Pectinatites. 
This  is  the  apertural  modification  of  the  microconch.  Most  microconchs  of  Ataxio- 
ceras appear  to  have  well-developed  lappets  (e.g.  Geyer  1961a,  pi.  14,  fig.  2),  but  there 
are  specimens  which  appear  to  have  a  horn  developed  (e.g.  Geyer  1961a,  pi.  13,  fig.  5). 
It  thus  appears  that  three  of  the  most  important  characters  of  Pectinatites  are  also 
found  in  Ataxioceras. 

There  must  also  be  taken  into  account  the  remarkable  similarity  of  some  species 
of  Pectinatites  to  species  of  the  Tithonian  genera  Subplanites,  Lithacoceras,  and 
Pseudovirgatites.  There  can  be  no  doubt  that  there  was  a  marine  connection  between 
Britain  and  the  Swabia-Franconia  area  at  least  for  a  short  while  after  the  Lower 
Kimmeridgian,  since  the  genus  Gravesia  is  common  to  both  areas.  However,  there 
are  apparently  no  substantiated  records  of  Lithacoceras  or  Subplanites  from  the 
Lower  Kimmeridgian  of  Britain.  (In  this  respect  I  cannot  accept  Arkell's  report  of 
Lithacoceras  from  the  Aulacostephanus  zones  (1947  :  73);  or  that  of  Ziegler  of  Sub- 
planites rueppellianus  from  the  same  beds  (1962  :  13)).  All  reported  instances  of 
these  genera  must,  in  the  absence  of  any  figured  evidence  to  the  contrary,  be  inter- 
preted as  occurrences  of  hitherto  undescribed  perisphinctids  which  in  the  author's 
opinion  do  not  belong  either  to  Subplanites  or  Lithacoceras. 

The  difference  between  Subplanites  and  Lithacoceras  on  the  one  hand,  and  Pectina- 
tites sensu  lato  on  the  other  hand  has  not  been  recognized  hitherto  owing  to  the  failure 
to  take  note  of  the  different  types  of  dimorphism  in  the  two  faunal  provinces.  As 
shown  above,  the  microconchs  of  Pectinatites  are  horned,  the  macroconchs  have  a 
straight  peristome  and  often  a  tendency  towards  virgatotome  ribbing  on  the  body- 
chamber. 

Subplanites  and  Lithacoceras  both  have  lappeted  microconchs,  so  that  microconchs 
with  intact  peristomes  are  easily  distinguishable  from  microconchs  of  Pectinatites. 
However,  several  species  of  microconchs  of  Subplanites  such  as  S.  reisi,  S.  schlosseri, 
and  5.  moernsheimensis  have  polygyrate  ribbing  on  their  body-chamber,  and  are 
similar  in  adult  size  and  rib-style  to  macroconchs  of  species  of  Pectinatites,  but  differ 
in  that  the  former  bear  lappets  whereas  the  latter  do  not.  When  material  without 
intact  peristomes  is  compared,  therefore,  the  two  forms  are  virtually  indistinguish- 
able. Similarly  Lithacoceras  can  be  confused  with  Pectinatites  when  peristomes  are 
not  intact.  Thus,  previous  comparisons  of  the  British  Kimmeridgian  fauna  to  the 
Tithonian  fauna  have  been  comparisons  between  Kimmeridgian  macroconchs  and 
Tithonian  microconchs.  The  similarity  of  the  two  faunas  therefore  must  be  regarded 
as  an  example  of  penecontemporaneous  homeomorphy. 

However  the  similarity  of  the  rib-style  of  the  two  faunas,  in  particular  the  presence 


24  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

of  polygyrate  ribbing,  strongly  suggests  that  they  were  derived  from  the  same  stock. 
This  presumably  lay  in  some  of  the  less  specialized  of  the  Lower  Kimmeridgian 
ataxioceratids,  or  in  such  a  genus  as  the  Upper  Oxfordian  Discosphinctes  which  has 
some  polygyrate  ribs  on  the  body-chamber. 

This  being  the  case  it  would  seem  that  Lithacoceras  should  be  classified  together 
with  Subplanites  and  Pectinatites  in  the  same  sub-family.  Arkell  (1957)  placed 
Lithacoceras  in  the  sub-family  Ataxioceratinae  (Buckman  1921)  whilst  Subplanites 
and  Pectinatites  were  assigned  by  him  to  the  sub-family  Virgatosphinctinae  (Spath 
1923).  Since  these  three  genera  are  presumed  to  be  derivatives  of  the  ataxioceratid 
stock,  and  not  themselves  ataxioceratids,  they  perhaps  should  be  all  placed  together 
in  the  sub-family  Virgatosphinctinae. 

Subgenus  ARKELLITES  nov. 

Type  species.     Pectinatites  (Arkellites)  hudlestoni  sp.  nov. 

Diagnosis.  Dimorphic.  Microconchs  fairly  coarsely  ribbed  on  inner  whorls. 
Body-chamber  generally  more  coarsely  ribbed  than  inner  whorls.  Horn  sometimes 
little  more  than  an  inflation  of  ventral  part  of  peristome.  Macroconchs  with 
similarly  ribbed  inner  whorls  to  those  of  microconchs.  Outer  whorl  showing  little 
or  no  variocostation,  ribs  little  changed  to  the  peristome.  Some  species  showing 
strengthening  of  primary  ribs  with  development  of  intercalatory  secondary  and 
unbranched  primary  ribs.  Polygyrate  ribs  generally  rare.  Peristome  simple. 
Constrictions  if  present  shallow.     Suture  line  unknown. 

Upper  Kimmeridgian,  Elegans  to  Hudlestoni  Zones,  ?  Lower  Pectinatus  Zone. 

Pectinatites  {Arkellites)  primitivus  sp.  nov. 

(PL  2,  figs.  1,  2;   PI.  3) 

Diagnosis.  Macroconchs  125-150  mm.  in  diameter,  with  following  rib  densities: 
at  15  mm.  30-32  ribs;  at  20,  32-34;  25,  33-35;  30,  33-36;  35,  36-37;  40,  37-38; 
45 »  37-38;  50,39;  55,40;  60,41;  65,43.  Ribs  rectiradiate  to  prorsiradiate  with 
wide  angle  of  furcation.  Outer  whorl  variable  but  typically  with  frequent  un- 
branched primary  ribs.  Microconchs  80-105  mm-  in  diameter,  with  following  rib 
densities:  at  15  mm.  32  ribs;  at  20,  32-36;  25,  33-37;  30,  34-38;  35,  35-38;  40,  36-40. 
Ribs  rectiradiate  or  slightly  prorsiradiate.  Outer  whorl  variable  with  bifurcate  ribs 
predominating,  sometimes  with  polygyrate  and  simple  ribs,  and  intercalatory 
secondaries.     Peristome  slightly  inflated  ventrally. 

Holotype.     Macroconch  C. 73392. 

Paratype.     Macroconch  C. 73393. 

Paratypes  (allotypes).     Microconchs  C. 73394,  C. 73395. 

Material.     Nine  specimens  (four  macroconchs,  five  microconchs). 

Horizon.  Holotype,  paratype  and  allotype  C .  73394,  from  25  ft.  below  the  Yellow 
Ledge  Stone  Band.     Allotype  C. 73395  from  55  ft.  below  this  band. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET 


25 


o 

a: 

UJ 

CD 

2 

D 


bU 

h 

H 

^ 

^■^ 

3U 

45 

D 

^ 

d 

30 

45 

"^J 

— "      C 



4U 

c 

30 

45 
40 

P 

P 

3D 

15 


25 


35  45 

DIAMETER  (MM) 


55 


65 


Fig.  3.  Rib  density  of  species  of  the  subgenus  Arkellites.  Upper  case  letters  :  macro- 
conchs  ;  lower  case  letters  :  microconchs.  H,  h  :  P.  (A.)  hudlestoni  ;  D,  d  :  P.  {A.) 
damoni  ;    C,  c  :    P.  [A.)  cuddlensis  ;    P,  p  :    P.  (A.)  primitivus. 


26  UPPER    KIMMERIDGE   CLAY    OF    DORSET 

Stratigraphical  range.  Upper  Kimmeridgian,  Elegans  Zone,  between  12 
and  55  ft.  below  the  Yellow  Ledge  Stone  Band  (see  below). 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  125-150  mm. 
Diameter  of  umbilicus  49-65  mm.  The  last  whorl  of  the  holotype  has  54  primary  and 
76  secondary  ribs,  while  the  last  whorl  of  the  paratype  has  approximately  48  primary 
and  90  secondary  ribs.  At  20  mm.  diameter  there  are  34  ribs;  at  25,  35;  30,  36; 
35.36;  40,37;  45.37-38;  50,39;  55.4°;  60,41;  65,43.  The  paratype  has  a 
similar  rib-density. 

The  ribs  on  the  inner  whorls  are  rectiradiate  at  the  umbilical  shoulder,  then  sweep 
forward  to  become  prorsiradiate.  (The  few  apparently  rursiradial  ribs  on  the 
holotype  have  been  distorted  by  crushing.)  The  point  of  bifurcation  of  the  ribs  is 
high  on  the  whorl  side.  The  umbilical  seam  gradually  uncoils  over  the  last  half- 
whorl.  The  ribs  on  the  outer  whorl  are  either  bifurcate  or  simple.  The  angle  of 
bifurcation  is  larger  than  in  most  of  the  species  of  this  subgenus.  The  holotype 
shows  no  fewer  than  19  simple  ribs  on  its  outer  whorl.  The  paratype  has  fewer 
simple  ribs  (approximately  9)  and  at  least  one  trifurcate  rib. 

Microconch.  Evolute  shell  with  a  diameter  of  between  80  and  105  mm.  Diameter 
of  umbilicus  32-40  mm.  The  last  whorl  of  paratype  C. 73395  has  approximately 
45  primary  and  103  secondary  ribs.  At  15  mm.  diameter  there  are  32  ribs,  at  20, 
32-36;  25,  33-37;    30,  34-38;    35.  35-38;   40,  36-40. 

The  ribs  on  the  inner  whorl  are  rectiradiate  at  the  umbilical  shoulder  but  curve 
forwards  and  become  straight  and  slightly  prorsiradiate  for  the  rest  of  their  length. 
The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl  side.  The  umbilical  seam 
uncoils  over  the  last  half  whorl.  The  ribs  on  the  outer  whorl  gradually  become 
straight  and  rectiradiate.  There  are  three  simple  ribs  on  the  last  whorl  of  paratype 
C. 73394,  otherwise  the  ribs  are  bifurcate.  The  other  paratype  shows  several  poly- 
gyrate  ribs  on  the  last  whorl.     The  peristome  is  slightly  inflated  ventrally. 

Remarks.  The  macroconch  of  this  species  shows  features  which  are  interpreted 
as  being  primitive  characters  of  the  genus.  These  include  the  wide  angle  of  furcation 
of  the  ribs,  the  abundant  unbranched  primary  ribs,  and  the  relative  absence  of 
trifurcate  (polygyrate)  ribs.  The  microconchs  exhibit  only  a  feeble  ventral  inflation 
of  the  peristome.  The  outer  whorl  of  allotype  C. 73395  shows  resemblance  in  its 
rib-style  to  microconchs  of  species  of  the  sub-genus  Virgatosphinctoides ,  but  the 
coarser  ribbing  of  the  inner  whorls  of  the  former  provides  easy  distinction.  It  is 
possible,  however,  that  the  subgenus  Virgatosphinctoides  was  derived  from  this  species. 

Incomplete  and  poorly  preserved  ammonites  from  the  lowest  Hen  Cliff  Shales, 
and  their  basal  cementstone,  may  belong  to  this  species.  This  would  extend  the 
range  of  the  species  down  to  70  feet  below  the  Yellow  Ledge  Stone  Band. 

Pectinatites  (Arkellites)  cuddlensis  sp.  nov. 

(PI.  4;   PL  5,  %•  1) 
Diagnosis.     Large    stoutly-ribbed    Arkellites.     Macroconchs    160-210    mm.    in 
diameter,  with  following  rib  densities:  at  30  mm.,  34-40  ribs;  at  35,  37-40;  40,  38-41; 


UPPER    KIMMERIDGE   CLAY    OF    DORSET  27 

45.  39-41 1  5°.  40_42;  55.  40-42;  60,  41-43;  65,  42-43.  Ribs  rectiradiate  to  slightly 
prorsiradiate.  Outer  whorl  with  mainly  bifurcate  ribs,  but  some  unbranched  primary 
and  occasional  polygyrate  ribs,  and  intercalatory  secondary  ribs.  Microconchs 
110-128  mm.  in  diameter,  with  following  rib  densities:  at  15  mm.,  33-34  ribs;  at  20, 
34-36;  25,  35-37;  30,  35-38;  35,  36-39;  4°-  36-41;  45,  37-  Ribs  of  inner  whorls 
similar  in  style  to  macroconch.     Peristome  with  ventral  horn  6-21  mm.  long. 

Holotype.     Macroconch  C. 73396. 

Paratype  (allotype).     Microconch  C. 73397. 

Material.     Nine  specimens  (five  macroconchs,  four  microconchs). 

Horizon.  Holotype  from  18  ft.  and  paratype  from  25  ft.  above  the  Yellow  Ledge 
Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  topmost  Elegans  to  lower 
Scitulus  Zones,  12  ft.  below  to  27  ft.  above  the  Yellow  Ledge  Stone  Band. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  160-210  mm.  The 
holotype  has  a  diameter  of  206  mm.,  and  an  umbilical  diameter  of  94  mm.  There  are 
49  primary  and  107  secondary  ribs  on  the  outer  whorl  of  the  holotype.  At  30  mm. 
diameter  the  holotype  has  34  ribs ;  at  35,  36;  40,38;  45,39;  50,40;  55,40;  60,41; 
65,42;  70,43;  75,44;  80,46;  85,47;  90,48-  The  variation  in  rib  density  within 
the  species  is  shown  in  Text-fig.  3. 

The  ribs  on  the  inner  whorls  are  rursiradiate  at  the  umbilical  shoulder;  they  then 
swing  forwards  and  become  rectiradiate  or  slightly  prorsiradiate  and  fairly  straight. 
Some  of  the  ribs  on  the  inner  whorls  of  the  holotype  are  partially  distorted  by  the 
crushing.  The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side.  There  is  a 
slight  uncoiling  of  the  umbilical  seam  over  the  last  half  whorl. 

The  ribs  on  the  outer  whorl  become  stronger  and  their  furcation  somewhat  irregular, 
with  the  development  of  intercalatory  secondary  ribs,  unbranched  primary  ribs, 
occasional  trifurcate  ribs,  and  occasional  furcation  low  in  the  whorl-side. 

There  do  not  appear  to  be  any  constrictions  on  the  last  whorl,  although  the  rib 
style  at  one  point  (a  trifurcate  rib  followed  very  closely  by  a  simple  rib)  is  very 
similar  to  that  which  obtains  when  a  constriction  is  present. 

The  peristome  is  simple. 

Microconch.  Evolute  shell  with  a  diameter  of  1 10-128  mm.  Diameter  of  um- 
bilicus 42-47  mm.  The  last  whorl  of  the  paratype  has  43  primary  and  89  secondary 
ribs.  At  15  mm.  diameter  the  paratype  has  33  ribs;  at  30,  34;  25,  36;  30,  38; 
35,  39;  4°,  41-  The  other  microconchs  of  this  species  are  similar  to  the  paratype  in 
rib  density.     (Text-fig.  3). 

The  ribs  on  the  inner  whorls  are  similar  in  style  to  those  of  the  macroconch.  The 
umbilical  seam  uncoils  over  the  last  half  whorl,  which  appears  to  correspond  to  the 
length  of  the  body-chamber.  The  ribs  on  the  outer  whorl  coarsen  slightly  and  are 
more  or  less  straight  and  rectiradiate.  There  are  occasional  simple  and  trifurcate 
ribs  on  the  last  whorl. 

The  aperture  bears  a  horn  which  is  21  mm.  long  on  the  paratype ;  the  other  complete 
microconchs  have  shorter  horns. 


28  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Remarks.  This  species  has  a  similar,  though  not  identical,  density  of  ribbing 
on  the  inner  whorls  to  that  of  P.  (A.)  primitivus,  described  above  (p.  24).  It  differs, 
however,  in  adult  size  of  both  macroconch  and  microconch,  and  the  density  and 
style  of  ribbing  of  the  outer  whorl.  It  is  probably,  nevertheless,  a  derivative  of  this 
former  species. 

Pectinatites  (Arkellites)  damoni  sp.  nov. 
(PI.  5,  figs.  2,  3;   PL  6) 

Diagnosis.  Macroconchs  136-160  mm.  in  diameter  with  following  rib  densities: 
at  25  mm.  39  ribs;  at  30,  40;  35,40-41;  40,41-42;  45,41-42;  50,42-43;  55,43-45- 
Ribs  rectiradiate  to  slightly  prorsiradiate.  Outer  whorl  with  irregular  ribs  with 
fairly  wide  angle  of  furcation,  occasional  unbranched  primary  ribs.  No  trifurcate 
ribs.  Microconchs  70-90  mm.  in  diameter  with  following  rib  densities:  at  15  mm. 
34-35  ribs;  at  20,  35-37;  25,  36-37;  3°,  36-37;  35.  37-3§-  Ribs  generally  pror- 
siradiate. Outer  whorl  with  occasional  unbranched  primary  and  polygyrate  ribs. 
Peristome  with  ventral  ribbed  horn  3-15  mm.  long. 

Holotype.     Macroconch  C. 73398. 

Paratype.     Macroconch  C. 73399. 

Paratypes  (allotypes).     Microconchs  C. 73400,  C. 73401. 

Material.     Sixteen  specimens  (four  macroconchs,  twelve  microconchs). 

Horizon.  Holotype  and  allotypes  from  25  ft.,  paratype  from  27  ft.  above  the 
Yellow  Ledge  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  lower  Scitulus  Zone,  ranging 
between  15  and  32  ft.  above  the  Yellow  Ledge  Stone  Band. 

Description.  Macroconch.  Moderately  evolute  shell  with  a  diameter  of  130-160 
mm.  Diameter  of  umbilicus  54-60  mm.  The  holotype  has  approximately  41 
primary  ribs  on  the  last  whorl.  No  macroconch  of  this  species  with  a  complete  last 
whorl  has  been  found.  At  25  mm.  diameter  the  holotype  has  39  ribs;  at  30,  40; 
35,40;  40,41;  45,41;  50,42;  55,43;  60,43.  The  variation  in  rib  density  within 
the  species  is  shown  in  Text-fig.  3. 

On  the  inner  whorls  the  ribs  are  rursiradiate  at  the  umbilical  shoulder,  they  then 
swing  forwards  to  become  rectiradiate  or  slightly  prorsiradiate  and  more  or  less 
straight.  The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side,  and  is  not 
always  visible  on  the  innermost  whorls.  There  is  a  slight  uncoiling  of  the  umbilical 
seam  over  the  last  half  to  three-quarters  of  a  whorl. 

The  ribs  on  the  outer  whorl  are  rather  "  untidy  "  in  appearance.  They  become 
coarser  and  have  quite  a  wide  angle  of  furcation.  Occasional  simple  ribs  are 
developed,  but  there  is  an  absence  of  trifurcate  ribs.     No  constrictions  are  visible. 

Although  no  specimen  has  its  peristome  preserved  intact,  it  is  presumably  simple. 

Microconch.  Evolute  shell  having  a  diameter  of  70-90  mm.  Diameter  of  um- 
bilicus 25-35  mm-  The  last  whorl  of  paratype  C. 73400  has  42  primary  and  86 
secondary  ribs.     The  density  of  the  ribs  on  the  inner  whorls  can  only  be  approxi- 


UPPER   KIMMERIDGE   CLAY    OF   DORSET  29 

mately  determined  on  this  specimen.  Paratype  C. 73401  has  35  ribs  at  both  25  and 
30  mm.  diameter.  Other  microconchs  of  the  species  show  similar  rib  densities 
(Text-fig.  3). 

The  ribs  of  the  inner  whorls  are  similar  in  style  to  those  of  the  macroconch,  but 
tend  to  be  more  prorsiradiate.  On  the  outer  whorl  the  forward  inclination  is  not  so 
pronounced,  and  the  ribs  approach  the  rectiradiate  condition.  The  point  of  bifurca- 
tion of  the  ribs  is  quite  high  on  the  whorl-side,  and  is  not  always  visible  on  the  inner 
whorls.  The  umbilical  seam  uncoils  gradually  over  the  last  half  whorl.  The  length 
of  the  body-chamber  (estimated  by  differences  in  the  degree  of  crushing)  is  usually 
half  a  whorl  long,  but  in  some  specimens  it  appears  to  be  only  about  three-eighths  of 
a  whorl  in  length. 

The  ribs  on  the  outer  whorl  are  slightly  coarser  than  those  of  the  inner  whorls. 
There  may  be  several  simple  and  occasional  trifurcate  ribs  on  the  last  whorl.  The 
ventral  part  of  the  peristome  bears  a  horn  which  varies  in  length  from  3-15  mm.  and 
is  always  quite  strongly  ribbed. 

Remarks.  The  macroconch  of  this  species  shows  some  similarity  with  that  of 
P.  {A.)  primitivus,  described  above  (p.  24).  It  differs  in  the  density  of  ribbing  of  the 
inner  whorls.  The  respective  microconchs  are  not  likely  to  be  confused  on  account  of 
size,  rib  style  and  density,  and  horn  development.  This  species  may,  however,  be 
derived  from  P.  {A.)  primitivus.  Adult  size  of  both  macroconch  and  microconch, 
and  the  very  irregular  ribbing  of  the  outer  whorl  distinguished  P.  (A.)  damoni  from 
P.  (A.)  cuddlensis. 


Pectinatites  (Arkellites)  hudlestoni  sp.  nov. 

(PI.  2,  fig.  3;   PI.  7;   PI.  8,  fig.  2) 

Diagnosis.  Large  Arkellites  with  stout  blunt  ribs.  Macroconchs  170-196  mm. 
in  diameter  with  following  rib  densities:  at  15  mm.  there  are  29-31  ribs;  at  20, 
30-34;  25,  31-34;  30,  32-35;  35,  34-37;  40,  36-39;  45,  36-40;  50,  38-43;  55, 
40-44;  60,  42-45.  Ribs  rectiradiate  to  slightly  rursiradiate.  Ribs  on  outer  whorl 
becoming  blunt  and  massive,  with  abundant  intercalatory  secondary  ribs  and 
occasional  unbranched  primary  ribs.  Microconchs  72-112  mm.  in  diameter,  with 
following  rib  densities:  at  15  mm.,  29-32  ribs;  at  20,  30-34;  25,31-36;  30,31-39; 
35.  33~42;  4°.  35~44!  45.  38-46.  Ribs  approximately  rectiradiate,  branching 
fairly  low  on  whorl-side.  Ribs  coarser  on  outer  whorl  with  occasional  simple  and 
rare  polygyrate  ribs.  Peristome  with  well-developed  ventral  horn  4-21  mm.  long. 
Horn  ornamented  only  by  growth  lines. 

Holotype.     Macroconch  C. 73403. 
Paratype  (allotype).     Microconch  C. 73404. 

Material.  Twenty-four  specimens,  including  eleven  plaster  casts.  Five  macro- 
conchs, nineteen  microconchs. 

Horizon.  Holotype  and  paratype  from  shales  T3  ft.  above  the  Rope  Lake  Head 
Stone  Band. 


3o  UPPER   KIMMERIDGE   CLAY   OF   DORSET 

Stratigraphical  range.  Upper  Kimmeridgian,  Hudlestoni  Zone,  from  12  ft. 
above  the  Rope  Lake  Head  Stone  Band  to  9  ft.  below  the  White  Stone  Band. 

Description.  Macroconch.  The  dimensions  given  below  are  from  the  holotype 
which  is  the  only  reasonably  complete  macroconch  which  is  well-preserved. 

Stoutly  ribbed  evolute  shell  with  a  diameter  of  196  mm.  Diameter  of  the  umbilicus 
93  mm.  There  are  42  primary  and  96  secondary  ribs  on  the  last  whorl.  At  15  mm. 
diameter  there  are  31  ribs;  at  20,  32;  25,32;  30,33;  35,34;  40,36;  45,36;  50, 
38;  55,4o;  60,42;  65,42;  70,43;  75,44;  80,45;  85,46;  90,46.  The  variation 
in  rib  density  is  shown  in  Text-fig.  3. 

The  ribs  on  the  inner  whorls  have  a  slight  rursiradial  curve  at  the  umbilical 
shoulder;  for  the  rest  of  their  length  they  are  straight  and  rectiradiate,  or  slightly 
rursiradiate.  The  point  of  bifurcation  is  high  on  the  whorl-side  apart  from  occasional 
ribs  which  branch  near  the  umbilical  shoulder.  The  umbilical  seam  uncoils  over  the 
last  five-eighths  of  a  whorl. 

On  the  outer  whorl  the  ribs  gradually  become  blunt  and  massive.  The  primary 
ribs  branch  to  give  only  two  secondary  ribs;  no  trifurcate  ribs  are  developed.  There 
are,  however,  abundant  intercalatory  secondary  ribs,  and  one  or  two  simple  ribs. 

The  peristome  is  not  preserved  completely,  but  in  the  absence  of  any  contrary 
evidence,  is  assumed  to  be  simple. 

Microconch.  Evolute  shell  with  a  diameter  of  72-112  mm.  Diameter  of  the 
umbilicus  26-45  mm.  The  paratype  has  a  diameter  of  98  mm.  and  an  umbilical 
diameter  of  44  mm.  It  has  55  primary  and  no  secondary  ribs  on  its  outer  whorl. 
At  15  mm.  diameter  the  paratype  has  31  ribs,  at  20,  33;  25,  35;  30,  36;  35,  39; 
40,  40;  45,  42.     The  variation  in  rib  density  is  shown  in  Text-fig.  3. 

The  ribs  on  the  inner  whorls  are  of  similar  style  to  those  of  the  macroconch,  but 
the  point  of  bifurcation  is  somewhat  lower  on  the  whorl-side.  The  umbilical  seam 
uncoils  over  the  last  half  whorl. 

The  ribs  on  the  last  whorl  gradually  become  coarser;  occasional  simple  ribs  are 
developed,  and  very  rarely  a  trifurcate  rib. 

The  peristome  is  straight  and  has  a  well  developed  ventral  horn.  On  the  paratype 
the  horn  projects  20  mm.  from  the  venter,  but  it  may  be  as  short  as  4  mm.  The  horn 
itself  is  ornamented  only  by  growth  lines,  the  ribs  on  the  whorl-side  fading  as  they 
approach  the  venter  in  the  vicinity  of  the  horn. 

Remarks.  This  species  is  younger  in  age  than  other  species  of  this  subgenus 
recorded  hitherto.  It  is  distinguished  by  its  rib-style,  particularly  the  tendency  for 
the  ribs  of  the  macroconch  to  be  slightly  rursiradiate.  The  microconch  horn  also 
tends  to  be  free  of  ornamentation.  On  some  specimens  (e.g.  C. 73402  figured  in 
PI.  2,  fig.  3)  the  venter  bears  scars,  suggesting  that  earlier  formed  horns  may  have 
been  shed. 


Subgenus  VIRGATOSPHINCTOIDES  Neaverson  1925  :  n 
1925     Allovirgatites  Neaverson  :  29. 
Type  species.     Virgatosphinctoid.es  wheatleyensis  Neaverson  1925. 


UPPER    KIMMERIDGE   CLAY    OF   DORSET  31 

Diagnosis.  Dimorphic.  Macroconchs  generally  finely  ribbed  on  inner  whorls. 
Outer  whorl  very  variable;  primary  ribs  strong  and  typically  with  frequent  poly- 
gyrate  furcation,  and  often  a  tendency  to  become  fasciculate  or  virgatotome ;  second- 
ary ribs  obsolescent  in  some  large  species.  Variocostation  slight  to  pronounced. 
Peristome  simple.  Microconchs  similarly  ribbed  on  inner  whorls  to  macroconchs. 
Body-chamber  usually  more  coarsely  ribbed  than  inner  whorls.  Peristome  typically 
with  well-developed  ventral  horn,  more  rarely  only  with  ventrally  inflated  peristome. 
Constrictions  commonly  present,  particularly  in  macroconchs. 

Upper  Kimmeridgian,  Elegans  to  Hudlestoni  Zones. 


Pectinatites  (Virgatosphinctoides)  elegans  sp.  nov. 
(PL  8,  figs.  la,  lb;   PL  9) 

Diagnosis.  Macroconchs  154-184  mm.  in  diameter  with  following  rib  densities: 
at  15  mm.  diameter  there  are  approximately  37  ribs;  at  20,  38;  25,  39;  30,  39-42; 
35,  40-43;  40,  4i-45;  45,  42-46;  50,  43-47;  55,  44-48;  60,  45-49;  65,  46-50. 
Ribs  slightly  prorsiradiate,  fairly  straight.  Outer  whorl  developing  strengthened 
primary  ribs,  mainly  bifurcate,  but  with  occasional  simple  and  polygyrate  ribs  and 
intercalatory  secondaries.  Microconchs  100- 112  mm.  in  diameter  with  following 
rib  densities:  at  20  mm.  diameter  there  are  39-44  ribs;  at  25,  40-45;  30,  41-45;  35, 
42-46;  40,  44-47.  Ribs  of  inner  whorls  similar  to  macroconch,  outer  whorl  slightly 
more  strongly  ribbed  with  occasional  simple  and  polygyrate  ribs.  Peristome  with 
ventral  inflation  projecting  4-8  mm. 

Holotype.     Macroconch  C. 73405. 

Paratype  (allotype).     Microconch  C. 73406. 

Material.     Fifteen  specimens  (six  macroconchs,  nine  microconchs). 

Horizon.  Holotype  from  18  ft.  and  paratype  from  20  ft.  below  the  Yellow 
Ledge  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  Elegans  Zone,  between  50  and 
16  ft.  below  the  Yellow  Ledge  Stone  Band. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  154-184  mm. 
Diameter  of  the  umbilicus  65-76  mm.  There  are  61  primary  and  137  secondary 
ribs  on  the  last  whorl  of  the  holotype.  At  40  mm.  diameter  the  holotype  has  45  ribs, 
at  45,  66;  50,47;  55,48;  60,48;  65,49;  7°,  5°-  The  variation  in  rib  density  is 
shown  in  Text-fig.  4. 

The  sharp  dense  ribs  on  the  inner  whorls  are  slightly  prorsiradiate  and  straight 
for  most  of  their  length,  but  at  the  umbilical  shoulder  are  rectiradiate  or  rursiradiate 
as  with  other  species  of  this  subgenus.  The  point  of  bifurcation  of  the  ribs  is  very 
high  on  the  whorl  side.     The  umbilical  seam  uncoils  over  the  last  half -whorl. 

The  ribs  on  the  outer  whorl  are  similar  in  style  to  those  of  the  inner  whorls  but 
gradually  become  more  blunt  and  more  widely  spaced.  Occasional  simple  and 
polygyrate  ribs  are  developed  on  the  outer  whorl,  but  bifurcate  ribs  predominate. 

The  peristome  is  simple. 


32 


UPPER  KIMMERIDGE  CLAY  OF  DORSET 


bO 

00 

W 

40 

1 

___ ' 

30 

w 

45 

S 

D 

/ 

35 

__d__ — 

5D 

s 

S 

40 

50 

40 

e 

E 

35 

15 


25 


35 


45 


55 


65 


DIAMETER  (MM.) 

Fig.  4.  Rib  density  of  species  of  the  subgenus  Virgatosphinctoides .  Upper  case  letters  : 
macroconchs  ;  lower  case  letters  :  microconchs.  W,  w  :  P.  (V.)  woodwavdi  ;  D,  d  :  P. 
(V.)  decorosus  ;    S,  s  :    P.  (V.)  scitulus  ;    E,  e  :    P.  (V.)  elegans. 


UPPER   KIMMERIDGE   CLAY   OF   DORSET  33 

Microconch.  Evolute  shell  with  a  diameter  of  100-112  mm.  Diameter  of  the 
umbilicus  (paratype)  40  mm.  The  paratype  has  66  primary  and  approximately 
130  secondary  ribs  on  the  last  whorl.  At  15  mm.  diameter  the  paratype  has  approxi- 
mately 42  ribs;  at  20,  44;  25,45;  30,45;  35,46;  40,47.  The  variation  in  rib  density 
is  shown  in  Text-fig.  4. 

The  rib  style  of  the  inner  whorls  is  very  similar  to  that  of  the  macroconch.  The 
umbilical  seam  uncoils  over  the  last  half  whorl.  No  suture  is  visible,  but  differences 
in  the  degree  of  crushing  suggest  that  the  body-chamber  is  half  a  whorl  in  length. 
The  ribs  on  the  last  half  whorl  become  slightly  coarser,  and  occasional  simple  and 
trifurcate  ribs  are  developed.  The  peristome  curves  forward  ventrally  and  is 
inflated  on  the  ventral  margin  which  projects  8  mm.  on  the  paratype.  The  ribs 
pass  uninterrupted  over  the  projection. 

Remarks.  This  is  the  earliest  known  species  of  the  subgenus  Virgatosphinctoides. 
The  origin  of  the  subgenus  may  have  been  from  P.  (Arkellites)  primitivus  described 
above  (p.  24),  the  microconch  of  which  shows  a  broadly  similar  rib  style  to  this  species, 
It  is  readily  distinguished,  however,  by  the  much  more  finely  ribbed  inner  whorls 
and  the  pronounced  peristomal  inflation.  P.  (V.)  elegans  is  distinguished  from  later 
species  of  the  subgenus  by  its  rib-density,  the  peristomal  development  of  the  micro- 
conch and  the  body-chamber  ornament  of  the  macroconch. 

Pectinatites  {Virgatosphinctoides)  elegans  corniger  subsp.  nov. 

(PL  10) 
Diagnosis.  Macroconch  approximately  125  mm.  in  diameter  with  following 
approximate  rib  densities:  at  30  mm.  48  ribs;  at  35,  49;  40,  50;  45,  51;  50,  52. 
Ribs  prorsiradiate,  fairly  straight.  Outer  whorl  developing  strengthened  primary 
ribs  with  fairly  frequent  polygyrate  ribs.  Microconch  82-85  mm.  in  diameter  with 
following  rib  densities :  at  20  mm.  diameter  approximately  43  ribs;  at  25,  45;  30,46. 
Inner  whorls  similarly  ribbed  to  macroconch,  outer  whorl  slightly  more  coarsely 
ribbed  with  occasional  simple  and  polygyrate  ribs.  Peristome  with  ventral  horn 
up  to  7  mm.  long. 

Holotype.     Macroconch,  C. 73407. 

Paratypes  (allotypes).     Microconchs.     C. 73408,  C. 73409. 

Material.     The  holotype  and  two  paratypes. 

Horizon.  Holotype  and  paratype  C. 73408  from  5  ft.,  and  paratype  C. 73409 
from  8  ft.  below  the  Yellow  Ledge  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  topmost  Elegans  Zone,  between 
5  and  8  ft.  below  the  Yellow  Ledge  Stone  Band  (see  below). 

Description.  Macroconch.  Evolute  shell  with  an  estimated  diameter  of  123  mm. 
Diameter  of  the  umbilicus  50  mm.  The  last  whorl  has  approximately  53  primary 
ribs.  The  number  of  secondary  ribs  cannot  be  determined  since  a  part  of  the  outer 
whorl  has  been  broken  away.  At  39  mm.  diameter  there  are  48  ribs;  at  35,  49; 
40.  5o;   45.  5i;   50,  52. 

GEOL.  15,  I.  3 


34  UPPER    KIMMERIDGE   CLAY   OF   DORSET 

The  ribs  on  the  inner  whorls  are  of  a  similar  style  to  those  of  P.  {A.)  elegans 
described  above  but  tend  to  be  a  little  more  prorsiradiate.  The  point  of  bifurcation 
of  the  ribs  is  high  on  the  whorl  side.  The  umbilical  seam  uncoils  over  the  last  half 
whorl.  The  last  umbilical  whorl  shows  the  development  of  occasional  trifurcate  and 
simple  ribs;  these  become  more  numerous  on  the  outer  whork,  where  there  is  some 
degree  of  variocostation.  Over  the  last  half  whorl  the  primary  ribs  become  more 
widely  spaced  and  more  pronounced,  and  there  are  frequent  polygyrate  ribs. 

The  peristome  is  not  completely  preserved,  but  is  presumably  simple. 

Microconch.  Evolute  shell  with  a  diameter  of  82-85  mm-  Diameter  of  the 
umbilicus  28-32  mm.  There  are  55-63  primary  and  122-130  secondary  ribs  on  the 
last  whorl.     At  20  mm.  diameter  there  are  43  ribs;   at  25,  45;   30,  46. 

The  rib  style  of  the  inner  whorls  is  similar  to  that  of  the  macroconch.  The  umbili- 
cal seam  uncoils  over  the  last  half  whorl,  which  appears  to  correspond  to  the  length 
of  the  body-chamber,  to  judge  by  differences  in  the  degree  of  crushing. 

The  ribs  on  the  outer  whorl  are  of  similar  style  to  those  of  the  inner  whorls,  but 
become  slightly  coarser  with  the  tendency  to  develop  occasional  polygyrate  and 
simple  ribs.  The  peristome  is  curved  forwards  dorsally,  and  ventrally  has  a  horn 
which  is  7  mm.  long  on  paratype  C. 73408.  The  secondary  ribs  pass  uninterrupted 
over  the  horn. 

Remarks.  One  fragment  of  an  ammonite  possibly  belong  to  this  subspecies  was 
collected  from  the  Yellow  Tedge  Stone  Band,  thus  the  stratigraphical  range  of  the 
subspecies  may  extend  upwards  to  the  top  of  the  Elegans  Zone. 

The  subspecies  is  intermediate  in  many  respects  between  P.  (V.)  elegans  (p.  31)  and 
P.  (V.)  scitulus  (p.  34).  It  is  intermediate  in  age  between  the  two,  and  shows  charac- 
ters of  both  species.  It  is  distinguished  from  the  former  species  by  the  smaller  adult 
size  of  both  its  macroconch  and  microconch,  the  peristomal  development  of  the 
microconch,  and  the  more  strongly  ribbed  body  chamber  of  the  macroconch.  Dis- 
tinction from  the  latter  species  is  based  on  the  adult  size  of  the  macroconch  and 
microconch,  the  perstomal  development  of  the  microconch,  and  the  ribbing  on  the 
body-chamber  of  the  macroconch  which  is  not  so  markedly  variocostate  as  in  P.  (V.) 
scitulus.  The  rib-density  of  the  subspecies  shows  a  closer  relationship  to  P.  (V.) 
scitulus  than  to  P.  (V.)  elegans;  however,  the  subspecies  is  assigned  to  the  latter 
species  because  the  ornament  of  the  body-chamber  of  the  macroconch,  and  the  ribbed 
horn  of  the  microconch  show  more  affinity  to  the  developments  of  these  characters  in 
P.  (V.)  elegans.  This  subspecies  appears  to  form  a  direct  phylogenetic  link  between 
P.  (V.)  elegans  and  P.  (V.)  scitulus. 


Pectinatites  (Virgatosphinctoides)  scitulus  sp.  nov. 
(PI.  11) 

Diagnosis.  Macroconchs  130-162  mm.  in  diameter  with  following  rib  densities: 
at  30  mm.  diameter  approximately  48  ribs;  at  35,  46-49;  40,46-49;  45,47-49;  50, 
48-49;  55,49-50;  60,49-50;  65,49-51.  Ribs  rectiradiate  to  slightly  prorsiradiate, 
straight.     Outer  whorl  with  strengthened  primary  ribs  and  variable  number  of 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  35 

simple  and  polygyrate  ribs  with  occasional  intercalatory  secondary  ribs.  Micro- 
conchs  67-82  mm.  in  diameter  with  following  rib  densities:  at  20  mm.  diameter 
42-44  ribs;  at  25,  44-46;  30,45-47;  35,48.  Inner  whorls  similar  to  macroconch. 
Outer  whorl  with  slightly  stronger  ribs  with  tendency  to  be  flexuous,  and  with 
occasional  simple  and  polygyrate  ribs.  Peristome  with  feebly-ribbed  ventral  horn 
7-16  mm.  long. 

Holotype.     Macroconch  C.73411. 

Paratype  (allotypes).     Microconchs  C. 73412,  C. 73413. 

Material.     Twenty-two  specimens  (ten  macroconchs,  twelve  microconchs). 

Horizon.  Holotype  from  24  ft.  above  the  Yellow  Ledge  Stone  Band.  Paratypes 
from  25  and  15  ft.  respectively  above  this  band. 

Stratigraphical  range.  Upper  Kimmeridgian,  lower  half  of  Scitulus  Zone, 
occurring  in  the  Yellow  Ledge  Stone  Band  and  up  to  44  ft.  above  this  horizon. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  130-162  mm.  The 
holotype  is  162  mm.  diameter.     Diameter  of  the  umbilicus  50-78  mm.     There  are 

47  primary  and  99  secondary  ribs  on  the  last  whorl  of  the  holotype,  other  specimens 
have  similar  rib  density  on  the  last  whorl.     At  30  mm.  diameter  the  holotype  has 

48  ribs;  at  35,  49;  40,49;  45,49!  5o,  49;  55,  5°;  60,50;  65,51;  70,52;  75,52. 
The  variation  in  the  rib  density  is  shown  in  Text-fig.  4. 

The  ribs  on  the  inner  whorls  are  dense  and  sharp.  At  the  umbilical  shoulder  there 
is  a  slight  rursiradial  curve;  the  ribs  then  swing  forwards  and  become  straight  and 
rectiradiate  or  slightly  prorsiradiate.  The  point  of  bifurcation  of  the  ribs  is  high  on 
the  whorl-side.     The  umbilical  seam  uncoils  over  the  last  half  whorl. 

The  outer  whorl  develops  coarser  ribs;  the  primary  ribs  become  more  widely 
spaced,  and  there  is  the  development  of  simple  and  polygyrate  ribs,  and  intercalatory 
secondary  ribs.  There  appear  to  be  several  constrictions  of  the  shell  over  the  last 
whorl,  but  the  crushing  does  not  allow  this  to  be  definitely  ascertained.  There  is 
considerable  variability  in  the  sculpture  of  the  outer  whorl,  some  specimens  having 
very  frequent  polygyrate  ribs.  The  peristome  is  simple.  The  length  of  the  body- 
chamber  is  unknown. 

Microconch.  Evolute  shell  with  a  diameter  of  67-82  mm.  Diameter  of  the 
umbilicus  25-32  mm.  Paratype  C. 73413  has  45  primary  and  92  secondary  ribs  on 
the  last  whorl;  other  specimens  have  similar  rib  density  on  the  last  whorl.  At 
20  mm.  diameter  there  are  42  ribs;  at  25,  44;  30,45.  Variation  in  rib  density  of  the 
inner  whorls  is  shown  in  Text-fig.  4. 

The  ribs  on  the  inner  whorls  are  identical  in  style  to  those  of  the  macroconch.  The 
umbilical  seam  uncoils  over  the  last  half  whorl.  Differential  crushing  suggest 
that  the  body-chamber  varies  in  length  from  half  to  five-eighths  of  a  whorl. 

The  outer  whorl  has  somewhat  coarser  ribs  which  tend  to  be  a  little  flexuous.  At 
the  umbilical  shoulder  they  are  rursiradiate,  then  swing  forwards  to  become  slightly 
prorsiradiate,  rectiradiate,  or  slightly  rursiradiate.  There  are  occasional  simple  and 
polygyrate  ribs  on  the  last  whorl,  which  may  have  a  few  constrictions. 


36  UPPER   KIMMERIDGE   CLAY    OF   DORSET 

The  peristome  has  a  ventral  horn  which  varies  in  length  from  7  to  16  mm.  The 
horn  is  only  feebly  ribbed. 

Remarks.  The  points  of  distinction  between  this  species  and  P.  (V.)  elegans  and 
P.  (V.)  elegans  corniger  have  been  discussed  above  (p.  34).  It  is  readily  distinguished 
from  P.  (V.)  decorosus  described  below  (p.  36)  by  the  rib  density  of  the  inner  whorls. 


Pectinatites  (Virgatosphinctoides)  decorosus  sp.  nov. 

(PI.  12) 

Diagnosis.  Macroconchs  120-140  mm.  in  diameter  with  following  rib  densities: 
at  30  mm.  diameter  there  are  41  ribs;  at  35,  41;  40,  41-42;  45,  41-42;  50,  41-42. 
Ribs  of  inner  whorls  fairly  straight  and  prorsiradiate ;  outer  whorl  developing  coarser 
more  widely  spaced  rectiradiate  ribs,  with  occasional  constrictions  followed  by 
simple  unbranched  primary  rib.  Microconchs  approximately  85  mm.  in  diameter 
with  following  approximate  rib  densities:  at  25  mm.  diameter  there  are  38  ribs; 
at  30,  39;  35,  42.  Inner  whorls  ribbed  similarly  to  macroconch,  outer  whorl  with 
stronger  rectiradiate  ribs,  occasionally  simple  or  polygyrate.  Peristome  with  ventral 
horn  up  to  9  mm.  long. 

Holotype.     Macroconch  C. 73414. 

Paratype  (allotype).     C.  73415. 

Material.     Eight  specimens  (four  macroconchs,  four  microconchs). 

Horizon.     Holotype  and  paratype  from  15  ft.  above  the  Yellow  Ledge  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  Scitulus  Zone,  15  to  30  ft.  above 
the  Yellow  Ledge  Stone  Band. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  120-140  mm. 
Diameter  of  the  umbilicus  54-66  mm.  The  last  whorl  of  the  holotype  has  43  primary 
and  approximately  96  secondary  ribs.  At  30  mm.  diameter  there  are  41  ribs;  at 
35,41;  40,41-42;  45,41-42;   50,41-42. 

The  ribs  on  the  innermost  whorls  are  very  slender  and  delicate;  they  are  fairly 
straight  and  slightly  prorsiradiate.  An  initial  rursiradial  curve  then  develops  at 
the  umbilical  shoulder ;  the  ribs  then  swing  forwards  to  become  prorsiradiate.  The 
point  of  bifurcation  of  the  ribs  is  quite  high  on  the  whorl  side.  The  umbilical  seam 
uncoils  over  the  last  half  whorl. 

The  ribs  on  the  outer  whorl  gradually  become  coarser  and  more  widely  spaced, 
and  lose  most  of  their  initial  rursiradial  curve  to  become  fairly  straight  throughout 
their  length.  They  are  mainly  rectiradiate,  but  vary  from  slightly  rursiradiate  to 
slightly  prorsiradiate.  There  appear  to  be  four  or  five  constrictions  on  the  outer 
whorl  of  the  holotype.  These  are  usually  preceded  by  a  polyploke  rib,  formed  by 
the  fusion  close  to  the  umbilical  shoulder  of  two  bifurcate  ribs,  and  are  always  followed 
by  a  simple  rib.  Apart  from  these  modifications,  and  the  occasional  intercalatory 
rib,  all  the  ribs  are  bifurcate. 

The  peristome  is  simple. 


UPPER    KIMMERIDGE   CLAY    OF    DORSET  37 

Microconch.  E volute  shell  with  a  diameter  of  approximately  85  mm.  All  the 
measurements  given  are  from  the  paratype  which  is  the  only  well-preserved  micro- 
conch. Diameter  of  the  umbilicus  35  mm.  The  last  whorl  has  46  primary  and 
approximately  go  secondary  ribs.     At  25  mm.  diameter  there  are  38  ribs;   at  30,  39; 

35.  42. 

The  rib  style  on  the  inner  whorls  is  similar  to  that  of  the  macroconch.  The  um- 
bilical seam  uncoils  over  the  last  half  whorl  (which  is  estimated  to  be  the  length  of 
the  body-chamber).  The  ribs  on  the  outer  whorl  are  strong  and  rectiradiate;  they 
are  mainly  bifurcate,  but  there  are  occasional  simple  and  polygyrate  ribs. 

The  peristome  has  a  ventral  horn  of  moderate  length  (9  mm.  on  the  paratype) 
and  which  is  ribbed.  It  arises  gradually  as  an  extension  of  the  venter,  and  does  not 
project  very  sharply  as  is  the  case  in  several  other  species  of  the  genus. 

Remarks.  The  density  and  style  of  ribbing  of  this  species  render  it  readily 
distinguishable  from  allied  species  such  as  P.  (V.)  scitulus  described  above  (p.  34). 
The  rib  density  of  the  inner  whorls  shows  some  similarity  to  that  of  P.  (A.)  cuddlensis 
described  above  (p.  26),  but  adult  size,  rib  style,  and  rib  density  of  the  outer  whorl 
differ  markedly  in  the  two  forms. 

Pectinatites  (Virgatosphinctoides)  major  sp.  nov. 

(PI-  13) 

Diagnosis.  Very  large  Virgatosphinctoides  with  little  varicocostation.  Diameter 
approximately  240-320  mm.,  with  following  rib  densities:  at  50  mm.  diameter  there 
are  42-47  ribs;  at  60,  44-48;  70,  46-50;  80,  48-51;  90,  49-53;  100,  50-55;  no, 
51-55;  120,52-56;  130,54-56;  140,57.  Ribs  of  inner  whorls  slender,  rectiradiate 
to  prorsiradiate ;  outer  whorl  with  stronger  rectiradiate  ribs,  with  some  simple  and 
polygyrate  ribs. 

Holotype.     Macroconch  C. 73410. 

Material.     Ten  specimens  (all  macroconchs). 

Horizon.     Holotype  from  6  ft.  below  the  Yellow  Ledge  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  Upper  part  of  Elegans  Zone 
and  Scitulus  Zone,  20  ft.  below  the  Yellow  Ledge  Stone  Band  up  to  the  Cattle  Ledge 
Stone  Band. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  240-320  mm. 
The  diameter  of  the  umbilicus  varies  from  130-150  mm.  The  holotype  is  318  mm. 
in  diameter,  has  an  umbilical  diameter  of  140  mm.,  and  59  primary  and  124  secondary 
ribs  on  the  last  whorl.  At  50  mm.  diameter  it  has  46  ribs;  at  55,  47;  60,48;  65,49; 
70.50;  75.50;  80,51;  85,52;  90,53;  95,54;  100,55;  105,55;  110,55;  rr5,55; 
120,56;    125,65;    130,56;    135,57;    I40,57- 

The  ribs  on  the  inner  whorls  are  slightly  rursiradiate  at  the  umbilical  shoulder, 
and  then  bend  forwards  and  become  either  rectiradiate  or  slightly  prorsiradiate. 
The  point  of  bifurcation  of  the  ribs  occurs  fairly  high  on  the  whorl-side.  The  um- 
bilical seam  uncoils  over  the  last  half  whorl.     The  length  of  the  body-chamber  is 


38 


UPPER    KIMMERIDGE   CLAY   OF   DORSET 


9U 

\  G 

70 

/     P 

60 

50 

45 

M 

55 
50 

L 

40 
^5 

20 


40 


60  80 

DIAMETER  (MM) 


100 


120 


Fig.  5.     Rib  density  of  species  of  the  subgenus  Virgatosphinctoides .     G  :   P.  (V.)  grandis  ; 
P:    P.  (V.)  pseudoscruposus  ;    M:    P.  (V.)  major;    L:    P.  (V.)  laticostatus . 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  39 

unknown,  but  would  appear  to  be  greater  than  two-thirds  of  a  whorl,  if  the  differential 
crushing  is  a  reliable  guide.  The  ribs  on  the  outer  whorl  gradually  lose  their  initial 
rursiradial  curve,  and  become  rectiradiate  throughout.  There  is  little  variocostation 
in  this  species,  and  the  ribs  of  the  outer  whorl  remain  predominantly  bifurcate. 
There  are,  however,  occasional  simple  and  more  rarely  polygyrate  ribs.  The  holo- 
type  shows,  in  addition,  one  rib  near  the  smooth  peristome  margin,  which  bifurcates 
very  low  on  the  whorl-side  and  again  higher  on  the  whorl,  producing  a  total  of  four 
secondary  ribs.  This  type  of  furcation  (polyploke)  is  not  seen  on  any  other  specimen 
of  the  species,  otherwise  the  other  specimens  show  little  variation  from  the  holotype, 
except  that  some  of  the  specimens  from  higher  horizons  tend  to  develop  a  slightly 
lower  point  of  furcation  of  the  ribs  on  the  body-chamber. 

The  microconch  of  this  species  is  unknown. 

Remarks.  The  very  large  size  of  this  species  renders  it  readily  distinguishable 
from  other  species  of  Pectinatites  of  the  same  age.  It  may  be  distinguished  from 
other  large  species  of  the  genus  (all  of  which,  described  hitherto,  are  of  younger  age) 
by  the  very  small  degree  of  variocostation. 


Pectinatites  (Virgatosphinctoides)  clavelli  sp.  nov. 

(PL  14) 

Diagnosis.  Macroconchs  210-260  mm.  in  diameter  with  following  approximate 
rib  densities :  at  40  mm.  diameter  there  are  46  ribs;  at  50,49;  60,54-57;  70,55-61; 
80,  56-66.  Ribs  of  inner  whorls  slender  and  rectiradiate.  Outer  whorl  developing 
widely-spaced  massive  blunt  primary  ribs  with  frequent  polygyrate  furcation,  and 
intercalatory  secondary  ribs.  Microconchs  67-87  mm.  diameter  with  following 
rib  densities:  at  25  mm.  diameter  there  are  42  ribs;  at  30,  43-44;  35,  44.  Inner 
whorls  similarly  ribbed  to  macroconch.  Outer  whorls  somewhat  more  coarsely- 
ribbed  with  occasional  polygyrate  and  simple  ribs.  Peristome  projecting  ventrally 
by  up  to  5  mm. 

Holotype.     Macroconch  C. 73432. 

Paratypes  (allotypes).     Two  microconchs,  C. 73433,  C. 73434. 

Material.  Eleven  specimens,  all  plaster  casts  (five  macroconchs,  six  micro- 
conchs). 

Horizon.  Holotype  from  8  ft.,  and  paratypes  from  3  ft.  above  the  Grey  Ledge 
Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  lower  Wheatleyensis  Zone, 
between  3  and  28  ft.  above  the  Grey  Ledge  Stone  Band. 

Description.  Macroconch.  Large  evolute  shell  with  a  diameter  of  approximately 
210-260  mm.     The  diameter  of  the  umbilicus  varies  from  104  to  130  mm. 

The  last  whorl  of  the  holotype  which  is  212  mm.  in  diameter  has  51  primary  and 
an  estimated  122  secondary  ribs.  The  innermost  whorls  are  not  completely  preserved 
in  any  one  specimen.  At  60  mm.  diameter  the  holotype  has  57  ribs,  at  70,  62 ;  80,  66 ; 
90,  69;   100,  71.     The  variation  in  rib  density  is  shown  in  Text-fig.  6. 


4° 


UPPER    KIMMERIDGE    CLAY    OF    DORSET 


70 

60 

W 

' 

C 

w 

50 

-""""s 

c 

40 
35 

s , 

60 

m 

50 

E^- 

e 

40 

/_ 

45 
40 

A 

15 


25 


35 


45 


55 


65 


DIAMETER  (MM) 
Fig.  6.     Rib  density  of  species  of  the  subgenus  Virgatosphinctoides.     Upper  case  letters  : 
macroconchs  ;    lower  case  letters  :    microconchs.     W,  w  :    P.  (V.)  wheatleyensis  ;    C,  c  : 
P.  (V.)  clavelli  ;   S,  s  :  P.  (V.)  smedmorensis  ;   m  :   P.  (V.)  magnimasculus  ;   E,  e  :  P.  {V.) 
encombensis  ;    A  :   P.  (V.)  abbreviatus , 


UPPER    KIMMERIDGE    CLAY    OF   DORSET  41 

The  ribs  on  the  inner  whorl  are  rursiradiate  at  the  umbilical  shoulder;  they  then 
swing  forwards  and  become  straight  and  more  or  less  rectiradiate.  The  point  of 
bifurcation  of  the  ribs  is  high  on  the  whorl-side.  There  is  a  marked  uncoiling  of 
the  umbilical  seam  over  the  last  half  whorl. 

The  ribs  on  the  outer  whorl  become  coarser,  and  the  primary  ribs  become  more 
widely  spaced  and  irregular  in  their  style  and  furcation.  Some  are  rursiradiate, 
others  rectiradiate  or  prorsiradiate.  Several  of  the  primary  ribs  show  polygyrate 
furcation,  and  there  is  also  a  profusion  of  simple  and  intercalatory  ribs  on  the  last 
whorl. 

The  peristome  is  simple. 

Microconch.  Evolute  shell  considerably  smaller  than  the  macroconch,  having  a 
diameter  of  only  68-87  mm.  The  diameter  of  the  umbilicus  is  30-37  mm.  There 
are  46  primary  and  94  secondary  ribs  on  the  last  whorl  of  paratype  C  .73434.  The 
rib  style  of  the  inner  whorls  is  similar  to  that  of  the  macroconch.  At  25  mm.  diameter 
there  are  42  ribs ;  at  30,43-44;  at  35,  44.     (Text-fig.  6). 

The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side.  The  umbilical  seam 
uncoils  noticeably  over  the  last  half  whorl. 

The  outer  whorls  of  both  specimens  are  rather  distorted  by  the  crushing  so  that 
it  is  not  easy  to  deterine  the  original  rib  direction.  It  would  appear,  however,  to 
be  rursiradiate  at  the  umbilical  shoulder,  then  becoming  straight  and  rectiradiate. 
There  are  occasional  simple  ribs  on  the  last  whorl  and  a  few  primary  ribs  with  three, 
or  in  one  case,  four  secondary  ribs. 

The  ventral  part  of  the  perstome  bears  a  short  ribbed  horn  4-5  mm.  in  length. 
It  arises  gradually  from  the  venter  and  is  not  well  preserved  on  either  paratype. 

Remarks.  The  adult  diameter  of  the  macroconch  and  microconch,  their  rib 
density,  and  the  development  of  the  microconch  horn,  serve  to  distinguish  this 
species  from  others.  P.  (V.)  smedmorensis  described  below  is  considerably  more 
coarsely  ribbed  than  this  species.  As  the  highest  beds  of  the  underlying  Scitulus 
Zone  have  hitherto  yielded  no  ammonites,  it  is  not  possible  to  determine  the  origin 
of  this  species.  The  degree  of  variocostation  of  the  macroconch  is  more  pronounced 
than  that  of  P.  (V.)  scitulus,  but  the  general  rib  style  of  the  two  species  shows  some 
similarities.  The  very  irregular  costation  of  the  body-chamber  of  the  macroconch 
is  a  feature  characteristic  of  many  of  the  younger  species  of  the  subgenus  Virgato- 
sphinctoides.  The  development  of  the  microconch  horn  in  this  species  is  not,  however, 
very  typical. 

Pectinatites  (Virgatosphinctoides)  smedmorensis  sp.  nov. 

(PL  15,  figs.  1,  2) 
Diagnosis.  Macroconchs  approximately  150  mm.  in  diameter  with  following 
approximate  rib  densities:  at  60  mm.  diameter  there  are  39-41  ribs;  at  65,  39-41; 
70,  40-42;  75,  41-43;  80,  42.  Ribs  on  inner  whorls  slender  and  prorsiradiate 
becoming  rursiradiate  with  strengthening  of  primary  ribs  on  body-chamber.  Poly- 
gyrate, simple  and  intercalatory  secondary  ribs  occur  occasionally.  Constrictions 
developed  over  last  two  whorls.     Microconchs  86-107  mm.  in  diameter  with  following 


42  UPPER   KIMMERIDGE   CLAY   OF   DORSET 

approximate  rib-densities:  at  20  mm.  diameter  there  are  34  ribs;  at  25,  25;  30,  27; 
35,  38.  Ribs  of  inner  whorls  similar  in  style  to  macroconch,  becoming  a  little  coarser 
on  outer  whorl.     Peristome  bearing  ventral  horn  up  to  8  mm.  long. 

Holotype.     Macroconch,  plaster  cast  C  73430. 

Paratype  (allotype).     Microconch,  plaster  case  C  73431. 

Material.  Five  specimens,  including  three  macroconchs  (two  of  which  are 
plaster  casts,  and  two  microconchs  (both  plaster  casts)). 

Horizon.     Both  type  specimens  are  from  22  ft.  below  the  Blackstone. 

Stratigraphical  range.  Upper  Kimmeridgian,  lower  Wheatleyensis  Zone, 
from  8  ft.  above  the  Grey  Ledge  Stone  Band,  to  22  ft.  below  the  Blackstone  (A 
vertical  range  of  41  ft.). 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  approximately 
150  mm.  Diameter  of  the  umbilicus  approximately  85  mm.  At  60  mm.  diameter 
the  holotype  has  39  ribs;  at  65,  39;   70,40;   75,41;   80,42.     (Text-fig.  6). 

The  ribs  on  the  inner  whorls  are  rectiradiate  at  the  umbilical  shoulder,  then  swing 
forwards  to  become  fairly  straight  and  prorsiradiate.  The  point  of  bifurcation  of 
the  ribs  is  fairly  high  on  the  whorl-side.  The  umbilical  seam  uncoils  over  the  last 
half  to  three-quarters  of  a  whorl. 

The  ribs  on  the  outer  whorl  gradually  lose  their  prorsiradiate  tendency  and  become 
straight  and  slightly  rursiradiate  throughout  their  length.  The  primary  ribs  become 
more  widely  spaced  and  very  sharp,  and  the  development  of  the  secondary  ribs 
becomes  irregular.     There  are  occasional  polygyrate,  simple  and  intercalatory  ribs. 

There  are  several  constrictions  present.  On  the  last  half- whorl  these  are  straight 
and  are  followed  by  a  simple  rib.  The  constrictions  developed  earlier  are  quite 
pronouncedly  prorsiradiate,  however.  These  oblique  constrictions  are  preceded  by 
a  trifurcate  rib  which  branches  low  on  the  whorl-side,  and  are  followed  by  a  simple 
rib. 

A  part  of  the  suture  line  is  present  on  one  specimen;  it  is  not  well-preserved, 
however,  but  does  show  stout  saddles  and  lobes.  Both  lateral  lobes  appear  to  be 
trifid. 

The  peristome  is  not  completely  preserved  on  any  specimen  but  is  presumably 
simple. 

Microconch.  Neither  microconch  is  particularly  well  or  completely  preserved. 
Both  specimens  are  plaster  casts.  The  paratype  is  quite  evolute  and  has  a  diameter 
of  86  mm.  The  diameter  of  the  umbilicus  is  36  mm.  There  are  an  estimated  46 
primary  ribs  on  the  last  whorl. 

The  ribs  of  the  inner  whorls  are  similar  in  style  to  those  of  the  macroconch.  At 
20  mm.  diameter  there  are  34  ribs;   at  25,  35;   30,37;   35,38.     (Text-fig.  6). 

The  outer  whorl  has  ribs  of  similar  style,  but  they  become  a  little  coarser.  There 
is  one  possible  constriction  present  at  the  aperture. 

The  ventral  part  of  the  peristome  is  damaged,  but  there  is  visible  the  basal  3  mm. 
of  a  horn,  the  original  length  of  which  may  have  been  7-8  mm. 

Remarks.     Most  characters  of  this  species  are  sufficiently  distinctive  to  separate 


UPPER    KIMMERIDGE   CLAY   OF   DORSET  43 

it  from  other  species.  It  is  distinguished  from  P.  (V.)  clavelli  described  above  (p.  39) 
by  the  smaller  adult  size  of  the  macroconch  and  the  more  coarsely-ribbed  inner 
whorls.  The  microconch  has  a  similar  rib-density  on  its  inner  whorls  to  that  of 
P.  (V.)  woodwardi  (p.  45),  but  may  be  distinguished  by  the  rib-style  of  both  inner 
and  outer  whorls. 


Pectinatites  {Virgatosphinctoides)  laticostatus  sp.  nov. 

(PI.  16) 

Diagnosis.  Large  Virgatosphinctoides  developing  massive  widely-spaced  primary 
ribs  over  last  two  whorls.  Diameter  230-320  mm.,  with  following  approximate 
rib  densities:  at  20  mm.  diameter  there  are  38  ribs;  at  30,  42;  40,45;  50,47;  60, 
48-52;  70,  49-55;  80,  49-55;  90,  47-56;  100,  45-53;  no,  43-51;  120,  40-49 
130,  38-45.  Ribs  of  inner  whorls  rectiradiate  to  prorsiradiate  becoming  more  widely 
spaced  from  ante-penultimate  whorl  onwards.  Outer  whorl  extremely  coarsely 
ribbed  with  abundant  intercalatory  secondary  ribs.     Microconch  unknown. 

Holotype.     Plaster  cast  C. 73416. 

Material.     Six  specimens,  including  two  plaster  casts  (all  macroconchs). 

Horizon.     Holotype  from  "  dicey  "  shales  19  ft.  below  the  Blackstone. 

Stratigraphical  range.  Upper  Kimmeridgian,  lower  Wheatleyensis  Zone 
ranging  from  3  ft.  above  the  Grey  Ledge  Stone  Band  to  13  ft.  6  in.  below  the  Black- 
stone  (a  vertical  range  of  47  ft.). 

Description.  Large  evolute  shell  with  a  diameter  of  232-320  mm.  Diameter  of 
umbilicus  130-180  mm.  The  holotype  has  27  primary  and  approximately  68 
secondary  ribs  on  the  last  whorl.  At  20  mm.  diameter  holotype  has  approximately 
38  ribs;  at  30,  42;  40,45;  50,47;  60,48;  70,49;  80,49;  9°.  47;  ioo.45,'  110,43; 
120,  40;   130,  38.     The  variation  in  rib  density  is  shown  in  Text-fig.  5. 

The  ribs  on  the  innermost  whorls  are  rursiradiate  at  the  umbilical  shoulder  then 
swing  forwards  to  become  rectiradiate  or  slightly  prorsiradiate.  A  coarsening  of 
the  ribs  develops  very  early,  and  the  last  three  whorls  become  progressively  more 
coarsely  ribbed.  There  is  a  gradual  loss  of  the  initial  rursiradial  curve  of  the  ribs 
and  they  become  straight  throughout  their  length.  The  point  of  bifurcation  of  the 
ribs  is  high  on  the  whorl-side,  and  the  angle  of  furcation  is  somewhat  larger  (at  least 
on  the  outer  whorls),  than  is  usual  in  this  subgenus.  The  umbilical  seam  uncoils 
over  the  last  half  whorl. 

The  outer  whorl  becomes  extremely  coarsely  ribbed  and  abundant  intercalatory 
secondary  ribs  are  developed. 

At  least  two  constrictions  are  present  on  the  holotype,  one  on  the  penultimate  and 
one  on  the  antepenultimate  whorl;  they  are  both  strongly  oblique.  In  each  case 
the  constriction  is  preceded  by  a  biplicate  rib,  which  branches  very  close  to  the 
umbilical  shoulder.     A  simple  rib  follows  the  constriction. 

The  peristome  is  not  preserved  intact  on  the  holotype,  but  is  presumed  to  be  simple. 

The  microconch  of  this  species  has  not  been  found  hitherto. 


44  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Remarks.  The  very  early  development  of  widely-spaced  ribs  in  this  species  is  an 
uncommon  character  in  this  subgenus,  and  is  therefore  a  very  useful  character  for 
identification  of  this  species.  The  problematical  Virgatosphinctoides  nodiferus 
Neaverson  (1925  :  14,  pi.  4,  fig.  1)  has  a  similar  style  of  ribbing  on  its  outer  whorl, 
but  its  rib-development  is  not  known  in  any  detail.  It  is  apparently  geologically 
younger  than  P.  (V.)  laticostatus. 

Pectinatites  {Virgatosphinctoides)  grandis  (Neaverson) 
(PI.  15,  fig.  3;   PI- 18) 

1925      Virgatosphinctoides  grandis  Neaverson  :  13,  pi.  4,  fig.  2. 
Material.     Eight  specimens;   seven  macroconchs,  one  possible  microconch. 

Stratigraphical  range.  Upper  Kimmeridgian,  upper  part  of  Wheatleyensis 
Zone,  between  3  and  17  ft.  below  the  Blackstone. 

Description.  Macroconch.  There  is  good  agreement  between  one  of  the  speci- 
mens here  figured  (PL  r8)  and  the  holotype,  which  came  from  Corton,  Dorset.  The 
former  has  a  diameter  of  365  mm.  The  umbilicus  has  a  diameter  of  168  mm.  There 
are  24  primary  and  approximately  82  secondary  ribs  on  the  last  whorl.  At  80  mm. 
diameter  there  are  86  ribs,  at  90,  87 ;  100,85;  110,85;  120,79;  x30.  77 '>  r4°>  71'. 
150,  68 ;  160,  64.  The  variation  in  rib-density  of  the  Kimmeridge  forms  is  shown 
in  Text-fig.  5. 

The  ribs  on  the  inner  whorls  are  rursiradiate  at  the  umbilical  shoulder  and  gradually 
swing  forwards,  so  that  less  than  half  way  up  the  whorl-side  they  become  prorsi- 
radiate.  On  the  last  umbilical  whorl  the  ribs  become  more  widely  spaced,  until  on 
the  outer  whorl  the  primary  ribs  are  very  strong  and  distant  from  one  another. 

There  are  often  large  numbers  of  secondary  ribs  to  each  primary  rib.  On  some 
specimens  there  are  regularly  as  many  as  five  secondary  ribs  to  each  primary  rib. 
The  ribs  become  straighter  on  the  last  part  of  the  body-chamber  and  slightly  prorsi- 
radiate  throughout  their  length.  The  secondary  ribs  tend  to  become  less  prominent 
and  several  primary  ribs  may  be  unbranched.  Some  intercalatory  secondary  ribs 
are  usually  present  on  the  last  whorl.  There  are  several  constrictions  present. 
They  are  preceded  by  a  compound  rib,  and  followed  by  a  simple  rib. 

The  peristome  is  presumably  simple. 

Microconch.  The  figured  microconch  comes  from  the  same  horizon  (17  ft.  below 
the  Blackstone)  as  the  earliest  recorded  macroconch  of  this  species. 

It  is  112  mm.  in  diameter.  The  diameter  of  the  umbilicus  is  45  mm.  There  are 
approximately  70  primary  ribs  on  the  last  whorl.  At  40  mm.  diameter  there  are 
approximately  68  ribs.  The  ribs  of  the  inner  whorl  are  similar  in  style  to  those  of 
macroconch.  The  outer  whorl  is  similarly  ribbed,  but  has  occasional  simple  and 
polygyrate  ribs  and  at  least  one  constriction.  The  aperture  bears  a  horn  which 
projects  from  the  venter  by  about  7  mm. 

Remarks.  There  is  a  great  disparity  in  size  between  the  microconch  (112  mm. 
diameter)  and  the  associated  macroconch  (approximately  280  mm.  diameter). 
However,  it  has  been  found  that  as  a  general  rule  the  microconch  is  usually  slightly 


UPPER   KIMMERIDGE   CLAY   OF   DORSET  45 

coarser-ribbed  than  its  macroconch  at  the  same  diameter.  In  this  case  the  micro- 
conch  has  68  ribs  at  45  mm.  diameter  while  the  macroconch  has  approximately  70  at 
this  diameter.  No  other  fine-ribbed  macroconchs  occur  at  this  horizon,  so  that  there 
can  be  little  doubt  that  this  specimen  is  the  microconch  of  P.  (V.)  grandis. 

The  size  of  this  species,  coupled  together  with  rib-style  and  density  distinguish  it 
from  other  species  of  the  genus. 


Pectinatites  ( Virgatosphinctoid.es)  grandis  acceleratus  subsp.  nov. 

(PL  19) 

Diagnosis.  Very  large  Virgatosphinctoides.  General  characters  similar  to  P. 
(V.)  grandis  (Neaverson)  but  development  of  modified  ornament  occurring  earlier. 
Ribs  of  outer  whorl  blunt  and  massive  with  few  secondaries.  Some  intercalatory 
secondary  ribs. 

Holotype.     Macroconch  C. 73422,  the  only  specimen. 

Horizon.  13  ft.  above  the  Rope  Lake  Head  Stone  Band.  (Upper  Kimmeridgian, 
basal  part  of  Hudlestoni  Zone). 

Description.  Large  evolute  shell  with  a  diameter  of  approximately  375  mm. 
Diameter  of  the  umbilicus  175  mm.  There  are  23  primary  and  approximately  60 
secondary  ribs  on  the  last  whorl. 

This  subspecies  is  similar  in  most  respects  to  P.  (V.)  grandis  described  above.  It 
differs  in  that  it  becomes  coarser-ribbed  earlier  in  development,  but  the  different 
types  of  sculpture  present  in  P.  (V.)  grandis  are  repeated  in  the  same  order,  but  at 
smaller  diameters. 

The  peristome  is  simple. 

Remarks.  This  subspecies  is  closely  related  to  P.  (V.)  grandis  and  must  be 
interpreted  as  a  direct  derivative  of  it.  There  is,  however,  a  thickness  of  approxi- 
mately 35  ft.  of  rock  between  the  highest  recorded  occurrence  of  P.  (V.)  grandis 
and  the  horizon  from  which  this  subspecies  came.  Most  of  the  intervening  rocks, 
however,  are  extremely  poorly  fossiliferous,  so  that  collection  failure  is  most  probably 
responsible  for  the  "  break  ". 

No  microconch  of  this  subspecies  has  been  found  hitherto. 

Pectinatites  (Virgatosphinctoides)  woodwardi  (Neaverson) 

(PI.  20) 

1925  Allovirgatites  woodwardi  Neaverson  :  31,  pi.  3,  fig.  1. 

1925  Allovirgatites  robustus  Neaverson  :  32,  pi.  3,  fig.  3. 

1925  Allovirgatites  versicostatus  Neaverson  :  32,  pi.  3,  fig.  4. 

1926  Allovirgatites  woodwardi  Neaverson  ;    Buckman,  pi.  637. 

Material.     Eleven  specimens  (five  macroconchs,  six  microconchs). 
Stratigraphical  range.     Upper  Kimmeridgian,  Wheatleyensis  Zone  (just  above 
the  middle),  ranging  between  15  and  9  ft.  below  the  Blackstone. 


46  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Description.  Macroconch.  Fairly  evolute  shell  with  a  diameter  of  150-185  mm. 
Diameter  of  umbilicus  74-88  mm.  The  last  whorl  of  the  specimen  here  figured 
which  is  approximately  155  mm.  diameter  has  an  estimated  46  primary  and  111 
secondary  ribs.  The  innermost  whorls  are  not  completely  preserved  in  any  specimen. 
At  50  mm.  diameter  there  are  49  ribs,  at  55,  52;  60,54;  65,55;  70,56.     (Text-fig.  4). 

The  ribs  on  the  inner  whorls  are  rursiradiate  at  the  umbilical  shoulder,  then  swing 
forwards  to  become  rectiradiate  or  slightly  prorsiradiate  and  more  or  less  straight. 
The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side. 

The  umbilical  seam  uncoils  over  the  last  half-whorl.  (This  uncoiling  is  not  notice- 
able in  the  plate  reproduced  herein,  owing  to  the  crushing  of  the  last  umbilical  whorl 
which  gives  an  incorrect  impression  of  the  amount  of  this  whorl  exposed.) 

The  ribs  on  the  outer  whorl  become  more  widely  spaced  and  stouter.  The  point 
of  furcation  is  sometimes  lower  on  the  whorl-side.  The  number  of  secondary  ribs 
per  primary  is  variable,  with  as  many  as  four  secondary  ribs  to  each  primary  rib. 
There  are  at  least  two  possible  constrictions  on  the  outer  whorl.  The  peristome  is 
not  preserved  intact  on  any  specimens,  but  is  presumably  simple. 

Microconch.  Fairly  evolute  shell  with  a  diameter  of  approximately  68-77  mm- 
The  diameter  of  the  umbilicus  is  25-30  mm.  The  figured  specimen  has  49  primary 
and  an  estimated  92  secondary  ribs  on  its  last  whorl.  At  20  mm.  diameter  there  are 
approximately  34  ribs;  at  25,  35;  30,36.  The  variation  in  rib  density  is  shown  in 
Text-fig.  4. 

The  inner  whorls  are  similar  in  rib  style  to  those  of  the  macroconch.  The  umbilical 
seam  uncoils  over  the  last  half  whorl  (not  well-shown  on  the  figured  microconch). 

The  ribs  of  the  outer  whorl  lose  most  of  their  initial  rursiradial  curve  and  are  almost 
straight  and  rectiradiate.  There  is  some  slight  variability  in  the  rib  direction, 
however,  from  slightly  rursiradiate  to  slightly  prorsiradiate.  There  are  occasional 
simple  and  trifurcate  ribs  on  the  last  whorl. 

The  peristome  is  not  preserved  intact  on  any  one  specimen.  The  figured  specimen 
shows  it  to  be  more  or  less  straight,  however.  The  ventral  part  of  the  peristome  on 
this  specimen  projects  about  2  mm.  and  is  then  broken,  so  that  it  is  safe  to  conclude 
that  a  horn  was  originally  present. 

Remarks.  The  Dorset  specimens  agree  closely  with  Neaverson's  figure  of  Allo- 
virgatites  woodwardi.  A .  robustus  Neaverson  is  merely  an  incomplete  specimen  of  the 
same  species,  apparently  a  little  thicker-whorled,  but  still  very  close  to  the  former 
species.  A .  versicostatus  Neaverson  is  also  very  close  to  this  species  and  may  possibly 
be  the  microconch.  The  differences  do  not  appear  to  be  sufficient  to  warrant  specific 
distinction.  The  association  of  P.  (V.)  woodwardi  with  P.  (V.)  wheatleyensis  (Neaver- 
son) is  also  indicative  of  the  similarity  of  the  Dorset  to  the  Oxford  material. 

Pectinatites  {Virgatosphinctoides)  wheatleyensis  Neaverson 

(PI.  21) 

1925      Virgatosphinctoides  wheatleyensis  Neaverson  :  12,  pi.  1,  fig.  1. 

1956     Subplanites   {Virgatosphinctoides)   wheatleyensis   (Neaverson)   Arkell  :  779,   pi.   40,   fig.    1. 

Material.     Eleven  specimens  (five  macroconchs,  six  microconchs). 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  47 

Stratigraphical  range.  Upper  Kimmeridgian,  Wheatleyensis  Zone  (just  above 
the  middle),  ranging  from  15  to  9  ft.  below  the  Blackstone. 

Description.  Macroconch.  One  specimen  from  Kimmeridge  here  figured, 
(PL  2i,  fig.  1)  agrees  extremely  closely  with  the  holotype  figured  by  Neaverson.  It 
has  a  diameter  of  132  mm.  The  diameter  of  the  umbilicus  is  57  mm.  There  are 
approximately  64  primary  and  156  secondary  ribs  on  the  last  whorl.  Another  speci- 
men, which  has  well  preserved  inner  whorls,  has  rib  density  as  follows:  at  25  mm. 
diameter  there  are  54  ribs,  30,  54;  35,  55;  40,  57;  45,  58;  50,  58;  55,  60.  The 
variation  in  rib  density  of  the  Kimmeridge  specimens  is  shown  in  Text-fig.  6. 

The  ribs  on  the  inner  whorls  are  rursiradiate  at  the  umbilical  shoulder,  then  swing 
forwards  to  become  rectiradiate  or  slightly  prorsiradiate.  The  point  of  bifurcation  of 
the  ribs  is  high  on  the  whorl-side.     The  umbilical  seam  uncoils  on  the  last  half  whorl. 

The  outer  whorl  is  very  variable.  The  primary  ribs  become  more  widely  spaced 
and  are  mostly  polygyrate  in  some  specimens.  Others  show  the  persistence  of  a 
more  conservative  type  of  ribbing,  with  more  bifurcate  than  polygyrate  ribs. 

One  specimen  shows  the  last  few  approximated  suture  lines,  but  as  these  are 
extremely  poorly  preserved,  comparison  with  the  suture  line  figured  by  Neaverson 
(1925,  text-fig.  B,  5)  is  not  possible.  However,  it  can  be  seen  from  these  suture  lines 
that  the  body-chamber  is  half  a  whorl  in  length. 

The  peristome  is  simple  and  straight. 

Microconch.  One  specimen  figured  herein  (C.  73426)  has  a  diameter  of  91  mm.  The 
diameter  of  the  umbilicus  (which  is  somewhat  elongated  by  crushing)  is  36  mm. 
There  are  56  primary  and  122  secondary  ribs  on  the  last  whorl.  The  inner  whorls 
are  badly  preserved,  so  that  it  is  not  possible  to  determine  the  rib  density  accurately. 
There  are,  however,  approximately  50  ribs  at  35  mm.  diameter.  The  other  figured 
specimen  (C.  73427)  has  at  15  mm.  diameter  47  ribs,  at  20,  47;  25,  48;  30,  49. 
(Text-fig.  6). 

The  ribs  on  the  inner  whorls  are  of  a  similar  style  to  the  macroconch.  The  umbili- 
cal seam  uncoils  over  the  last  half  whorl  (which  appears  to  correspond  to  the  length 
of  the  body-chamber). 

The  ribs  on  the  outer  whorl  become  slightly  coarser,  and  there  is  the  development 
of  occasional  polygyrate  and  simple  ribs. 

The  peristome  bears  a  horn  when  completely  preserved,  and  this  is  17  mm.  long 
on  specimen  C.  73426.     It  is  quite  strongly  ribbed. 

Remarks.  The  horizon  at  which  this  species  occurs  in  Dorset  is  much  lower 
than  that  quoted  by  Arkell  (1947  :  71).  Although  the  ammonites  from  just  below 
the  Basalt  Stone  Band  are  undoubtedly  somewhat  similar  in  appearance  to  this 
species,  the  outer  whorls  are  not  the  same.  The  associated  fauna  also  confirms  the 
identity  of  this  species.  Neaverson  placed  his  Wheatleyensis  Zone  immediately 
above  the  Gravesia  Zones,  which  is  too  low  in  the  succession. 

Pectinatites  (Virgatosphinctoides)  wheatleyensis  minor  subsp.  nov. 

(PL  24,  fig.  1) 
Diagnosis.     Macroconchs  small  (103-108  mm.  diameter)  with  following  approxi- 


48  UPPER   KIMMERIDGE   CLAY   OF   DORSET 

mate  rib  densities :  at  25  mm.  diameter  there  are  44  ribs;  at  30,  48;  35,46;  40,48; 
45,  49.  Ribs  of  inner  whorls  slender  and  approximately  rectiradiate.  Outer  whorl 
developing  strengthened  primary  ribs,  remaining  approximated,  with  polygyrate 
furcation  predominant. 

Holotype.     Macroconch  C. 73429. 

Material.     Two  specimens  (both  macroconchs). 

Horizon.  Both  specimens  from  17  ft.  below  the  Blackstone  (Upper 
Kimmeridgian,  middle  Wheatleyensis  Zone). 

Description.  Evolute  shell  with  a  diameter  of  103-108  mm.  (small  for  a  macro- 
conch). Diameter  of  umbilicus  43-45  mm.  There  are  approximately  53  primary  and 
144  secondary  ribs  on  the  last  whorl.  In  rib  style  this  subspecies  is  very  similar  to 
P.  (V.)  wheatleyensis,  but  is  somewhat  more  coarsely  ribbed.  At  20  mm.  diameter 
there  are  approximately  44  ribs,  at  30,  45;  35,  46;  40,  48;  45,  49.  The  point  of 
bifurcation  of  the  ribs  is  high  on  the  whorl-side.  The  umbilical  seam  uncoils  over 
the  last  half-whorl.  The  ribs  on  the  outer  whorl  are  identical  in  style  with  those  of 
the  holotype  of  P.  (V.)  wheatleyensis,  being  mostly  polygyrate. 

The  peristome  is  simple. 

Remarks.  The  microconch  of  this  subspecies  is  not  known.  Apart  from  the 
somewhat  more  coarsely  ribbed  inner  whorls,  and  smaller  adult  size,  this  subspecies 
is  similar  to  P.  (V.)  wheatleyensis.  Its  lower  stratigraphical  horizon  suggests  that 
it  is  a  possible  ancestor  of  this  species. 

Pectinatites  (Virgatosphinctoides)  wheatleyensis  delicatulus  (Neaverson) 

(PI.  27,  fig.  2) 

1925      Virgatosphinctoides  delicatulus  Neaverson  :  15,  pi.  1,  figs.  2  and  ?3. 

1925  Allovirgatites  tutcheri  Neaverson  :  30,  pi.  3,  fig.  2. 

1926  Allovirgatites  tutcheri  Neaverson  ;    Buckman,  pi.  692. 

Material.     Five  specimens  (four  macroconchs,  one  possible  microconch). 

Stratigraphical  range.  Upper  Kimmeridgian,  Wheatleyensis  Zone  (upper  part) 
between  7  and  4  ft.  below  the  Blackstone  (see  below). 

Description.  Macroconch.  The  Dorset  specimens  agree  closely  with  Neaver- 
son's  figure  of  the  holotype,  but  are  complete  individuals.  One  specimen  with  a 
diameter  of  approximately  130  mm.  and  an  umbilical  diameter  of  53  mm.  has  about 
78  primary  and  151  secondary  ribs  on  the  last  whorl.  The  ribs  on  the  inner  whorl 
are  a  little  finer  and  more  dense  than  those  of  P.  (V.)  wheatleyensis.  Some  specimens 
have  several  constrictions  of  the  last  umbilical  and  the  outer  whorl. 

The  main  point  of  difference  between  this  subspecies  and  P.  (V.)  wheatleyensis 
lies  in  the  more  finely  ribbed  outer  whorl,  and  the  constrictions  which  are  usually 
present.  The  constrictions  are  preceded  by  a  polygyrate  or  polyploke  rib  and  are 
followed  by  a  simple  rib.  In  some  cases  this  rib  is  prominent  like  the  flare  of  Lyto- 
ceras.     (As  is  the  case  with  Neaverson's  examples). 

The  peristome  is  simple.     The  body-chamber  is  a  half  whorl  in  length. 


UPPER    KIMMERIDGE   CLAY   OF   DORSET  49 

Microconch.  A  single  poorly  preserved  specimen  from  5  ft.  below  the  Blackstone 
is  possibly  the  microconch  of  this  subspecies.  It  is  91  mm.  in  diameter,  has  an 
umbilical  diameter  of  35  mm.  and  has  approximately  56  primary  and  112  secondary 
ribs  on  the  last  whorl.  The  inner  whorls  are  so  hidden  by  pyrite  aggregates,  however, 
that  it  cannot  be  with  certainty  referred  to  this  subspecies. 

The  umbilical  seam  uncoils  over  the  last  half  whorl.  The  peristome  is  missing, 
but  was  presumably  originally  horned. 

Remarks.  Allovirgatites  tutcheri  Neaverson  differs  from  V irgatosphinctoides 
delicatuliis  Neaverson  by  no  more  than  the  difference  between  two  individuals  of  the 
same  species.  The  general  similarity  to  P.  (V.)  wheatleyensis  justifies  separation  only 
at  subspecific  level. 

This  subspecies,  which  is  stratigraphically  a  little  younger  than  P.  (V.)  wheat- 
leyensis, is  certainly  a  derivative  of  it.  One  specimen  which  I  refer  to  P.  (V.) 
wheatleyensis  is  similarly  ribbed  to  this  subspecies  on  the  inner  whorls,  but  is  inter- 
mediate between  the  two  forms  in  the  ribbing  of  its  outer  whorl. 

Several  very  poorly  preserved  ammonites  from  the  shales  4—10  ft.  above  the 
Blackstone  are  provisionally  included  in  this  subspecies. 


Pectinatites  (Virgatosphinctoid.es)  pseudoscruposus  (Spath) 

(PI.  17) 

1936     Subplanites  pseudoscruposus  Spath  :  173,  fig.  2. 

1947     Subplanites  pseudoscruposus  Spath  ;    Arkell  :  77,  fig.  17,  1. 

Emended  diagnosis.  Diameter  of  shell  220-230  mm.  with  following  rib  densities : 
at  40  mm.  diameter  there  are  approximately  51  ribs;  at  50,  53-57;  60,  55-61;  70, 
60-66;  80,  66-71;  90,  70-76;  100,  75-78.  Ribs  on  inner  whorls  slender  and 
prorsiradiate.  Outer  whorl  developing  massive  blunt  primary  ribs  typically  with 
virgatotome  furcation,  with  up  to  six  secondary  ribs  to  each  primary. 

Material.     Six  specimens  (all  macroconchs). 

Stratigraphical  range.  Upper  Kimmeridgian,  Wheatleyensis  Zone  (upper 
part),  between  9  and  3  ft.  below  the  Blackstone. 

Description.  Targe  evolute  shell  with  a  diameter  of  220  to  approximately  230 
mm.  Diameter  of  umbilicus  1 10-^115  mm.  The  specimen  figured  herein  (C. 73418) 
has  36  primary  and  approximately  104  secondary  ribs  on  the  last  whorl  (about  one- 
eighth  of  a  whorl  is  missing) . 

At  50  mm.  diameter  this  specimen  has  57  ribs,  at  60,  59;  70,  66;  80,  72;  90,  76; 
100,  77;   no,  78.     The  variation  m  rib  density  is  shown  in  Text-fig.  5. 

The  ribs  on  the  inner  whorls  are  fine  and  slender.  They  are  curved  rursiradially 
at  the  umbilical  shoulder,  but  swing  forwards  rapidly  to  become  prorsiradiate  and 
almost  straight.  The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side.  The 
umbilical  seam  uncoils  over  the  last  half  to  three-quarters  of  a  whorl. 

The  sculpture  of  the  outer  whorl  is  extremely  variable,  no  two  specimens  being 
alike  in  this  respect.     Typically  there  are  developed  very  prominent  blunt  primary 

GEOL.   15,   I.  4 


5o  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

ribs  which  tend  to  become  virgatotome,  with  up  to  six  secondary  ribs.  There  are 
occasional  simple  and  intercalatory  secondary  ribs. 

Constrictions  are  also  present  on  the  last  whorl  of  some  specimens  of  this  species. 
These  are  preceded  by  a  virgatotome  rib  with  four  or  five  secondary  ribs,  and  are 
followed  by  a  simple  rib. 

The  microconch  of  this  species  is  unknown. 

Remarks.  The  validity  of  this  species  may  be  questioned,  as  it  does  not  entirely 
fulfil  the  requirements  of  Article  13  of  the  International  Code  of  Zoological  Nomen- 
clature. However,  the  outer  whorl  fragment  upon  which  Spath  based  this  species  is 
absolutely  characteristic  and  there  can  be  no  doubt  of  its  interpretation.  Spath's 
figure  is  misleading  in  that  the  inner  whorls  he  associated  with  the  outer  whorl 
fragment  almost  certainly  belong  to  a  different  species.  The  name  of  this  species  is 
well-known  to  those  familiar  with  British  Kimmeridgian  ammonites,  and  there  would 
seem  to  be  little  to  be  gained  by  rejecting  a  name  because  its  original  description  was 
legally  defective.  The  intention  here  has  been  to  give  a  more  satisfactory  definition 
of  the  species  by  figuring  a  more  complete  specimen  and  giving  a  comprehensive 
description. 

Pectinatites  (Virgatosphinctoides)  reisiformis  sp.  nov. 

(PI.  22;   PI.  23,  fig.  3) 

Diagnosis.  Macroconchs  155-255  mm.  in  diameter  with  following  rib  densities: 
at  20  mm.  diameter  there  are  48-52  ribs,  at  25,  49-54;  30,  52-56;  35,  54-58;  40 
55-58;  45,56-59;  5o,57-6i;  55.58-63;  60,59-64;  65,60-65;  70,61-66.  Ribs 
of  inner  whorls  slender  and  prorsiradiate.  Outer  whorl  becoming  suddenly  more 
coarsely-ribbed  with  polygyrate  then  virgatotome  ribs,  with  abundant  simple  and 
intercalatory  secondary  ribs.  Microconchs  78-110  mm.  in  diameter  with  following 
rib  densities:  at  20  mm.  diameter  there  are  approximately  42  ribs;  at  25,  42-44; 
30,  43-46;  35,  44-47;  40,  46-49;  45,  48-52.  Ribs  of  inner  whorls  similar  to 
macroconch.  Outer  whorl  more  coarsely  ribbed  with  occasional  simple  and  poly- 
gyrate ribs  and  intercalatory  secondaries.  Peristome  with  ventral  horn  10-20  mm 
long. 

Holotype.     Macroconch  C. 73435. 

Paratype  (allotype).     Microconch  C. 73436. 

Material.     Twenty-five  specimens  (seventeen  macroconchs,  eight  microconchs). 

Horizon.  Both  holotype  and  paratype  are  from  shales  13  ft.  above  the  Rope  Lake 
Head  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  uppermost  Wheatleyensis  and 
basal  Hudlestoni  Zones,  from  6  ft.  above  the  Blackstone  to  16  ft.  above  the  Rope 
Lake  Head  Stone  Band. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  155-255  mm. 
Diameter  of  umbilicus  65-118  mm.  The  holotype  has  a  diameter  of  174  mm.,  and 
an  umbilical  diameter  of  76  mm.     There  are  51  primary  and  118  secondary  ribs  on 


UPPER  KIMMERIDGE  CLAY  OF  DORSET 


5i 


eu 

70 

RD 

60 

^- 

intersex 

^ / 

,y 

rd 

50 

45 

*" 

60 

R 

/ 

50 

r 

40 

55 

D 

45 

-    d 

15 


25 


35  45 

DIAMETER   (MM.) 


55 


65 


Fig.  7.  Rib  density  of  species  of  the  subgenus  Virgatosphinctoid.es.  Upper  case  letters  : 
macroconchs  ;  lower  case  letters  :  microconchs.  RD,  rd  :  P.  (V.)  reisiformis  densi- 
costatus  ;    R,  r  :    P.  (V.)  reisiformis  ;    D,  d  :    P.  (V.)  donovani. 


52  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

the  last  whorl.  The  inner  whorls  are  well  preserved  on  the  holotype.  At  20  mm. 
diameter  there  are  48  ribs,  at  25,49;  30,52;  35,54;  40,  55;  45,56;  50,57;  55,58; 
60,  59;   65,  60;   70,  61 ;   75,  62.     The  variation  in  rib  density  is  shown  in  Text-fig.  7. 

The  ribs  on  the  inner  whorls  are  dense  and  slender.  They  are  rursiradiate  at  the 
umbilical  shoulder,  then  swing  forwards  to  become  prorsiradiate  (pronouncedly  so 
in  some  specimens).  The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side. 
The  umbilical  seam  uncoils  over  the  last  half  to  three-quarters  of  a  whorl. 

On  the  outer  whorl  the  ribs  become  rather  suddenly  more  widely  spaced.  At  first, 
polygyrate  ribs  are  developed;  then  as  the  primary  ribs  become  stouter  and  blunter, 
the  branching  has  a  tendency  towards  the  virgatotome  condition,  with  up  to  four 
secondary  ribs  to  each  primary  rib.  There  are  abundant  simple  and  intercalatory 
ribs  on  the  last  whorl. 

The  ribbing  of  the  outer  whorl  is  extremely  variable  in  this  species.  No  two  speci- 
mens are  alike  in  this  respect.  Some  of  the  larger  specimens  are  obviously  gerontic 
individuals,  the  last  secreted  half  whorl  or  so  of  their  shell  being  almost  devoid  of 
ornament. 

The  peristome  is  not  preserved  on  the  holotype,  but  other  specimens  show  it  to  be 
straight  and  simple. 

Microconch.  Evolute  shell  with  a  diameter  of  78-110  mm.  The  diameter  of  the 
umbilicus  varies  from  30-42  mm.  The  paratype  is  no  mm.  in  diameter  and  has  an 
umbilical  diameter  of  42  mm.  There  are  63  primary  and  126  secondary  ribs  on  the 
last  whorl  of  the  paratype.  At  25  mm.  diameter  the  paratype  has  44  ribs,  at  30,  45 ; 
35,  47;  40,  48.     The  variation  in  rib  density  is  shown  in  Text-fig.  7. 

The  ribs  of  the  inner  whorls  are  similar  in  style  to  those  of  the  macroconch,  but 
tend  not  to  be  so  markedly  prorsiradiate.  The  point  of  bifurcation  of  the  ribs  is  a 
little  lower  on  the  whorlside  than  on  the  macroconchs.  The  umbilical  seam  uncoils 
over  the  last  half  whorl. 

The  outer  whorl  becomes  more  coarsely  ribbed,  and  there  is  a  tendency  for  the  ribs 
to  become  somewhat  flexuous.  At,  or  just  below  the  point  of  furcation,  the  ribs 
bend  back  a  little.  This  feature  is  well  shown  on  the  paratype.  Almost  all  the  ribs 
on  the  outer  whorl  are  bifurcate,  but  there  are  very  occasional  simple,  polygyrate, 
and  intercalatory  ribs. 

The  peristome  is  somewhat  sinuous,  and  laterally  extends  anteriorly  a  little.  There 
is  a  well  developed  ventral  horn,  which  is  often  feebly  ribbed.  On  the  paratype  the 
horn  projects  from  the  venter  by  about  18  mm.  In  some  specimens  growth  has 
proceeded  a  little  beyond  the  horn;  this  anteriorly  extended  portion  of  the  shell 
shows  little  or  no  ornamentation. 

Remarks.  The  density  of  the  ribbing  on  the  inner  whorls,  and  the  sculpture  of 
the  outer  whorl  are  distinctive  features  of  this  species.  It  may  be  derived  from  P.  (V.) 
wheatleyensis  but  shows  considerably  more  variocostation  than  this  latter  species. 
The  diversity  of  the  ornamentation  of  the  body-chamber  of  the  macroconchs  con- 
trasts with  the  rib-density  of  the  inner  whorls  which  are  remarkably  similar  in  rib 
style  and  density. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  53 

Pectinatites  (Virgatosphinctoides)  reisiformis  densicostatus  subsp.  nov. 

(PI.  23,  figs.  1,  2;   PL  24,  fig.  2) 

Diagnosis.  Macroconchs  150-195  mm.  in  diameter  with  following  rib  densities: 
at  20  mm.  diameter  there  are  52-58  ribs;  at  25,  55-60;  30,  56-63;  35,  60-65;  4°. 
62-67;  45.  65-71;  50,  68-74;  55-  70-76;  60,  73-77;  65,  75-79-  Ribs  of  inner 
whorls  slender  and  prorsiradiate.  Outer  whorl  developing  strengthened  primary 
ribs,  first  with  polygyrate  and  polyploke  furcation,  then  with  tendency  to  virgatotome 
furcation.  Microconchs  77-110  mm.  diameter  with  following  rib  densities:  at  20 
mm.  diameter  there  are  47-50  ribs;  at  25,  48-52;  30,  50-54;  35,  51-55;  40,  51-56. 
Ribs  of  inner  whorls  slender  and  prorsiradiate,  becoming  somewhat  coarser  on  outer 
whorl  with  occasional  polygyrate  and  simple  ribs.  Peristome  with  ventral  horn 
12-24  mm.  long. 

Holotype.     Macroconch  C. 73437. 

Paratype.     Microconch  C. 73438. 

Material.  Twenty-eight  specimens  (fourteen  macroconchs,  thirteen  micro- 
conchs, one  intersex). 

Horizon.  Both  holotype  and  paratype  from  shales  13  ft.  above  the  Rope  Lake 
Head  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  lower  part  of  Hudlestoni  Zone, 
12-16  ft.  above  the  Rope  Lake  Head  Stone  Band. 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  150-195  mm. 
Diameter  of  the  umbilicus  63-87  mm.  The  holotype  is  157  mm.  in  diameter  and 
has  an  umbilical  diameter  of  67  mm  There  are  64  primary  and  141  secondary  ribs 
on  the  last  whorl.  At  15  mm.  diameter  there  are  about  52  ribs;  at  20,  58;  25,  60; 
30,63;  35,63;  40,65;  45,65;  50,68;  55,70;  60,74;  65,75.  The  variation  in 
rib  density  is  shown  in  Text-fig.  7. 

In  style  of  ribbing  on  both  inner  and  outer  whorls  this  subspecies  agrees  closely 
with  the  description  of  P.  (V.)  reisiformis  given  above  (p.  50).  It  differs,  however, 
in  being  far  more  densely  ribbed  throughout.  The  ribs  themselves  are  also  a  little 
more  slender. 

Microconch.  Evolute  shell  with  a  diameter  of  77-110  mm.  Diameter  of  the 
umbilicus  27-42  mm.  The  paratype  has  a  diameter  of  83  mm.  and  an  umbilical 
diameter  of  30  mm.  There  are  67  primary  and  132  secondary  ribs  on  the  last  whorl. 
At  15  mm.  the  paratype  has  an  estimated  46  ribs,  at  20,  50;  25,  50;  30,  52.  The 
variation  is  rib  density  is  shown  Text-fig.  7. 

The  rib  style  on  the  inner  whorls  is  similar  to  that  of  the  microconch  of  P.  (V.) 
reisiformis,  but  the  ribs  are  usually  a  little  more  prorsiradiate  and  are  more  slender. 

On  the  outer  whorl  the  ribs  become  a  little  coarser,  and  occasional  polygyrate  and 
simple  ribs  are  developed. 

The  peristome  margin  may  be  straight  or  sinuous,  a  horn  is  developed  ventrally, 
and  on  the  paratype  projects  from  the  venter  by  16  mm.  In  some  cases  growth  has 
proceeded  a  little  beyond  the  growth  of  the  horn,  and  there  is  a  fairly  smooth  zone 


54  UPPER   KIMMERIDGE   CLAY    OF   DORSET 

anterior  to  this.     In  other  cases  a  further  horn  may  be  grown  close  to  the  first  one. 

Intersex.  This  subspecies  is  particularly  interesting  because  of  a  probable  mutation 
which  arose  resulting  in  the  development  of  inter-sexual  individuals.  One 
specimen  (C. 73439,  PI.  24,  fig.  2)  is  intermediate  in  size  between  macroconch  and 
microconch  (117  mm.  diameter)  and  has  rib  density  of  a  typical  microconch  up  to  a 
diameter  of  30  mm.  (20,  47;  25,  49;  30,  52).  Thereafter  it  becomes  more  finely 
ribbed  (35,  56;  40,  58),  and  is  intermediate  between  macroconch  and  microconch  in 
rib  density.     (Text-fig.  7). 

At  a  diameter  of  94  mm.  a  horn  is  developed,  and  beyond  this  there  is  about  three- 
eighths  of  a  whorl  of  coarsely  ribbed  shell  with  sculpture  similar  to  the  outer  whorls 
of  a  macroconch,  but  bearing  four  further  horns.  The  development  in  this  subspecies 
of  the  macroconch  outer  whorl  sculpture  at  such  a  small  diameter  is  unique  to  this 
specimen. 

In  addition  to  the  above  specimen  which  is  absolutely  intermediate  in  character 
between  macroconch  and  microconch,  three  other  specimens  show  a  slight  degree  of 
intersexuality.  These  three  specimens  are  apparently  normal  macroconchs  to  judge 
by  their  size,  rib  density  and  sculpture.  They  do,  however,  develop  a  type  of  horn 
in  the  later  stages  of  development;  this  appears  at  a  diameter  of  140-150  mm.  and 
is  unlike  the  true  microconch  horn  in  that  it  is  developed  from  a  single  rib,  has 
negligible  ventral  projection  but  projects  laterally  some  distance  down  the  whorl 
side.  In  addition,  the  diameter  at  which  these  structures  are  developed  is  much 
greater  than  that  at  which  the  true  horn  of  the  microconch  occurs. 

These  three  latter  specimens  possess  some  degree  of  microconch  character. 

Remarks.  This  subspecies  shows  a  general  similarity  to  P.  (V.)  reisiformis 
described  above.  It  may  readily  be  distinguished,  however,  by  the  rib  density  of 
the  inner  whorls.  Because  of  this  likeness  to  P.  (V.)  reisiformis,  the  similar  horizon 
of  the  two  forms  (this  subspecies  is  confined  to  the  beds  yielding  the  youngest  speci- 
mens of  P.  (V .)  reisiformis)  tends  to  confirm  the  view  that  it  should  be  considered  a 
direct  derivative  of  P.  (V.)  reisiformis. 


Pectinatites  (Virgatosphinctoides)  abbreviatus  sp.  nov. 

(PL  26,  fig.  3) 
Diagnosis.  Macroconchs  small  (113-125  mm.  in  diameter)  with  following  rib 
densities:  at  20  mm.  diameter  there  are  40-42  ribs;  at  25,  41-43;  30,  42-44;  35, 
44-45 ;  40,  454-6 ;  45,  46.  Ribs  on  inner  whorls  slender,  rursiradiate  to  rectiradiate. 
Primary  ribs  becoming  more  widely  spaced  and  stronger  on  outer  whorl,  with  develop- 
ment of  simple  and  polygyrate  ribs.     Microconchs  unknown. 

Holotype.     Macroconch  C. 73440. 

Material.     Three  specimens  (all  macroconchs). 

Horizon.     Holotype  from  shales  20  ft.  above  the  Rope  Lake  Head  Stone  Band. 

Stratigraphical  range.     Upper  Kimmeridgian,  Hudlestoni  Zone  (lower  part), 
20.-22  ft.  above  the  Rope  Lake  Head  Stone  Band. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  55 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  1 13-125  mm. 
Diameter  of  umbilicus  45-59  mm.  The  holotype  has  a  diameter  of  113  mm.  and 
an  umbilical  diameter  of  45  mm.  There  are  47  primary  and  96  secondary  ribs  on  the 
last  whorl.  At  15  mm.  diameter  the  holotype  has  39  ribs ;  at  20,  40 ;  25,  41 ;  30,  42 ; 
35,  44;   40,  45;   45,  46.     The  variation  in  rib  density  is  shown  in  Text-fig.  6. 

The  ribs  on  the  inner  whorls  are  rursiradiate  at  the  umbilical  shoulder,  they  then 
bend  forwards  to  become  rectiradiate.  Some  of  the  ribs  are,  however,  rursiradiate 
throughout  their  length.  The  point  of  bifurcation  is  high  on  the  whorl-side.  The 
umbilical  seam  uncoils  over  the  last  half  whorl. 

On  the  outer  whorl,  the  ribs  gradually  become  coarser,  and  on  the  last  half  whorl 
(which  to  judge  by  differences  in  the  crushing  corresponds  to  the  length  of  the  body- 
chamber)  the  primary  ribs  become  widely  spaced.  Several  simple,  polygyrate  and 
intercalatory  ribs  are  developed. 

The  peristome  is  damaged  on  all  the  specimens,  but  on  the  holotype  it  is  partially 
preserved  and  appears  to  be  straight  and  simple. 

The  microconch  of  this  species  is  unknown. 

Remarks.  This  species  is  notable  for  the  small  size  at  which  the  macroconch 
becomes  mature.  For  this  reason  it  is  unlikely  to  be  confused  with  any  other  species 
of  the  genus.  P.  (V.)  wheatleyensis  minor  is  of  similar  size,  but  the  outer  whorl  of 
this  subspecies  is  ornamented  by  approximated  mainly  polygyrate  ribs ;  this  contrasts 
with  the  more  widely  spaced  primary  ribs  of  P.  (V.)  abbreviates. 


Pectinatites  (Virgatosphinctoides)  donovani  sp.  nov. 

(PL  25,  figs.  1,  2) 

Diagnosis.  Macroconchs  132-155  mm.  in  diameter  with  following  rib  densities: 
at  30  mm.  diameter  there  are  approximately  51  ribs;  at  35,  51;  40,  51-53;  45, 
51-54;  50,52-55;  55,53-55;  60,55-56.  Ribs  on  inner  whorls  slender  and  prorsi- 
radiate.  Outer  whorl  with  strong  irregular  primary  ribs  with  frequent  polygyrate 
furcation.  Simple  and  intercalatory  secondary  ribs  abundant.  Ribs  fading  slightly 
over  last  quarter  of  whorl.  Microconchs  99-106  mm.  in  diameter  with  following 
approximate  rib  densities:  at  30  mm.  diameter  there  are  46  ribs ;  at  35,  48;  40,49; 
45,  50.  Ribs  of  inner  whorls  similar  to  macroconch.  Outer  whorl  developing 
somewhat  stronger  ribs  with  occasional  simple  and  intercalatory  ribs  and  polygyrate 
furcation. 

Holotype.     Macroconch  C. 73441. 

Paratype  (allotype).     Microconch  C. 73442. 

Material.  Ten  specimens,  all  plaster  casts,  (six  macroconchs,  four  micro- 
conchs). 

Horizon.  Holotype  from  30  ft.  and  paratype  from  36  ft.  below  the  Basalt  Stone 
Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  Hudlestoni  Zone  (middle  part), 
30-40  ft.  below  the  Basalt  Stone  Band. 


56  UPPER    KIMMERIDGE   CLAY   OF   DORSET 

Description.  Macroconch.  Evolute  shell  with  a  diameter  of  132-155  mm. 
The  diameter  of  the  umbilicus  varies  from  59-69  mm.  The  holotype  has  a  diameter 
of  137  mm.  and  an  umbilical  diameter  of  65  mm.  The  last  whorl  of  the  holotype  has 
46  primary  and  101  secondary  ribs.  At  40  mm.  diameter  the  holotype  has  52  ribs, 
at  45,  53;  50,53;  55,54;  60,55;  65,56.  The  variation  in  rib  density  is  shown  in 
Text-fig.  7. 

The  ribs  on  the  inner  whorl  are  rectiradiate  at  the  umbilical  shoulder,  then  curve 
forwards  to  become  fairly  straight  and  prorsiradiate.  At  the  point  of  bifurcation, 
which  is  high  on  the  whorl-side,  the  ribs  curve  back  a  little  to  the  rectiradiate  position. 
The  umbilical  seam  uncoils  over  the  last  half  to  three-quarters  of  a  whorl. 

On  the  outer  whorl  the  ribs  gradually  lose  their  initial  rursiradial  curve.  The 
primary  ribs  become  very  strong  and  irregular,  and  the  point  of  furcation  is  lower  on 
the  whorl-side.  There  are  abundant  simple,  polygyrate  and  intercalatory  ribs  on  the 
last  whorl.  Over  the  last  quarter  of  a  whorl,  the  ribs  tend  to  fade  somewhat,  and 
although  easily  distinguishable,  are  not  so  prominent. 

The  peristome  is  somewhat  sinuous  and  is  simple. 

Microconch.  Evolute  shell  with  a  diameter  of  99-106  mm.  Diameter  of  the 
umbilicus  42-47  mm.  The  paratype  has  a  diameter  of  105  mm.  and  an  umbilical 
diameter  of  45  mm.  The  last  whorl  of  the  paratype  has  49  primary  and  96  secondary 
ribs.  At  30  mm.  diameter  the  paratype  has  46  ribs,  at  35,  48;  40,49;  45,50.  The 
variation  in  rib  density  is  shown  in  Text-fig.  7. 

The  ribs  on  the  inner  whorls  are  of  similar  style  to  those  of  the  macroconch.  The 
umbilical  seam  uncoils  over  the  last  half  whorl  (which  appears  to  correspond  to  the 
length  of  the  body-chamber,  to  judge  by  differences  in  the  degree  of  crushing). 

On  the  outer  whorl  the  ribs  become  rather  suddenly  more  widely  spaced  half  a 
whorl  from  the  aperture.  These  coarser  ribs  are  rather  irregular  in  their  furcation ; 
most  are  bifurcate,  but  there  are  several  simple,  polygyrate  and  intercalatory  ribs. 

The  peristome  is  not  completely  preserved  on  any  microconch  of  this  species  The 
dorsal  part  of  it  appears  to  be  straight,  but  no  specimen  shows  the  whole  of  the 
ventral  part  which  was  presumably  horned.  The  ventral  part  of  the  peristome  of 
the  paratype  appears  to  project  by  some  5  mm.,  but  is  not  well-preserved  at  this 
point. 

Remarks.  This  species  appears  to  be  the  one  misidentified  by  Arkell  as  P.  (V.) 
wheatleyensis  (Arkell  1956  :  21).  As  I  have  shown  earlier,  however  (p.  47),  the 
true  P.  (V.)  wheatleyensis  occurs  considerably  lower  in  the  succession,  where  it  is 
associated  with  other  species  also  characteristic  of  the  Nodule  Bed  of  Wheatley. 
The  rib  density  and  ornamentation  of  the  body-chamber  of  P.  (V.)  donovani 
distinguish  it  from  P.  (V.)  wheatleyensis. 

Pectinatites  (Virgatosphinctoides)  magnimasculus  sp.  nov. 

(PL  29) 
Diagnosis.     Microconchs  very  large  (175-185  mm.  in  diameter)  with  following 
approximate  rib  densities :   at  25  mm.  there  are  54  ribs;   at  30,  55;   35,57;   40,57; 
45,59;  50,60;  55,62;  60,62.     Ribs  of  inner  whorls  slender,  rectiradiate  to  slightly 


UPPER    KIMMERIDGE    CLAY    OF   DORSET  57 

prorsiradiate.     Outer  whorl  with  similar  style  of  ribs,  remaining  approximated  but 
becoming  a  little  coarser.     Peristome  with  ventral  horn  12-15  mm.  long. 

Holotype.     Microconch  C. 73443. 

Material.     Two  specimens,  both  plaster  casts,  microconchs. 

Horizon.     Holotype  from  21  ft.  below  the  White  Stone  Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  Hudlestoni  Zone  (upper  part), 
18-21  ft.  below  the  White  Stone  Band. 

Description.  Large  evolute  shell  with  a  diameter  of  175-185  mm.  Diameter 
of  umbilicus  77-82  mm.  The  number  of  primary  ribs  on  the  last  whorl  is  estimated 
as  about  70.  The  holotype  which  is  175  mm.  in  diameter  and  which  has  an  umbilical 
diameter  of  77  mm.  has  at  a  diameter  of  25  mm.  54  ribs;  at  30,  55;  35,  57;  40,  57; 
45.  59;  5°.  60;  55,  62;  60,  62.  (Text-fig.  6).  The  other  specimen's  inner  whorls 
are  too  poorly  preserved  to  measure  rib  density. 

The  ribs  on  the  inner  whorls  are  fine  and  slender.  They  are  rursiradiate  at  the 
umbilical  shoulder,  then  curve  forwards  to  become  straight  and  rectiradiate,  or 
slightly  prorsiradiate.  The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side. 
The  umbilical  seam  uncoils  over  the  last  half  whorl. 

The  ribbing  on  the  outer  whorl  is  similar  to  that  on  the  inner  whorls.  There  is 
gradual  coarsening  of  the  ribs  over  the  last  whorl,  but  they  are  still  approximated 
right  up  to  the  aperture. 

The  peristome  is  straight  and  bears  a  horn  15  mm.  long  on  the  holotype.  Two 
other  horns  are  also  visible  on  the  last  whorl  of  the  holotype.  These  are  11  and  13 
mm.  in  length,  and  are  ribbed.  The  other  microconch  specimen  has  three  horns 
which  in  order  of  age  are  6  +  ;   19  and  12+  mm.  in  length. 

Remarks.  Members  of  this  species  include  the  largest  known  horned  microconchs. 
It  is  interesting  to  note  that  the  ribs  on  the  last  whorl  differ  very  little  in  style  and 
density  from  those  of  the  earlier  whorls.  It  seems  probable  that  several  of  the  larger 
species  of  the  subgenus  Virgatosphinctoides,  of  which  no  microconchs  have  been 
found  hitherto,  have  microconchs  similar  in  size  to  those  of  P.  (V.)  magnismasculns. 

Associated  with  the  two  microconch  specimens  was  found  a  fragment  of  a  large 
macroconch,  which  if  complete  would  have  had  a  diameter  of  about  320  mm.  Its 
fragmentary  nature  and  extremely  poor  preservation  are  such,  however,  that  it 
cannot  be  referred  to  this  species  with  any  certainty. 

Pectinatites  (Virgatosphinctoides)  encombensis  sp.  nov. 

(PI.  27,  fig.  1;   PI.  28) 

Diagnosis.  Microconchs  70-103  mm.  in  diameter  with  following  rib  densities: 
at  20  mm.  diameter  there  are  42-43  ribs;  at  25,  43-47;  30,44-49;  35,46-52.  Ribs 
of  inner  whorls  very  slender,  rectiradiate  to  slightly  prorsiradiate.  Ribs  on  outer 
whorl  gradually  becoming  a  little  more  widely  spaced,  with  abundant  polygyrate 
furcation  on  body-chamber.  Peristome  with  ventral  horn  9-15  mm.  in  length. 
Macroconchs  155-215  mm.  in  diameter  with  following  very  approximate  rib  densities: 


58  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

at  55  mm.  diameter  there  are  52  ribs;  at  60,  54;  65,  56;  70,  57;  75,  58;  80,  59; 
85,  60;  go,  61;  95,  63;  100,  64;  105,  65;  no,  66;  115,  67.  Ribs  of  inner  whorls 
similar  to  microconch.  Outer  whorl  developing  strengthened  primary  ribs  becoming 
more  widely  spaced  with  occasional  simple  and  polygyrate  ribs  and  intercalatory 
secondaries. 

Holotype.     Microconch,  C. 73444. 

Paratype  (allotype).     Macroconch  C. 73445. 

Material.  Ten  specimens,  all  plaster  casts  (three  macroconchs,  seven  micro- 
conchs). 

Horizon.  Holotype  from  21  ft.  and  paratype  from  33  ft.  below  the  White  Stone 
Band. 

Stratigraphical  range.  Upper  Kimmeridgian,  Hudlestoni  Zone  (upper  part), 
9-33  ft.  below  the  White  Stone  Band. 

Description.  The  holotype  is  a  microconch,  since  no  really  adequately  preserved 
macroconchs  have  been  discovered. 

Microconch.  Evolute  shell  with  a  diameter  of  70-103  mm.  Diameter  of  the 
umbilicus  28-38  mm.  The  holotype  has  a  diameter  of  103  mm.  and  an  umbilical 
diameter  of  38  mm.  There  are  63  primary  and  approximately  133  secondary  ribs  on 
the  last  whorl  of  the  holotype. 

At  20  mm.  diameter  there  are  43  ribs,  at  25,  47,  30,48;  35,49.  The  variation  in 
rib  density  is  shown  in  Text-fig.  6. 

On  the  inner  whorls  the  ribs  are  very  slender.  They  are  rursiradiate  at  the  um- 
bilical shoulder,  then  bend  forwards  to  become  fairly  straight  and  rectiradiate  or 
slightly  prorsiradiate.     The  umbilical  seam  uncoils  over  the  last  half  whorl. 

The  outer  whorl  is  similarly  ribbed,  but  gradually  the  ribs  become  a  little  more 
widely  spaced  and  a  little  thicker.  Abundant  polygyrate  ribs  are  developed  on  the 
last  half  whorl  which  appears  to  correspond  to  the  length  of  the  body-chamber. 

The  peristome  is  fairly  straight  and  bears  a  ventral  horn  which  varies  in  length 
from  1 1-20  mm.  The  holotype  and  two  other  specimens  have  additional  horns  a 
little  way  back  from  the  peristome.  The  holotype  has  a  total  of  three  horns  which 
(in  order  of  age)  are  15,  15  plus  and  n  mm.  long. 

Macroconch.  Evolute  shell  with  a  diameter  of  155-215  mm.  Diameter  of 
umbilicus  75-118  mm.  The  paratype  is  the  only  macroconch  showing  any  detail 
of  the  inner  whorls.  It  is  215  mm.  in  diameter  and  has  an  umbilical  diameter  of 
118  mm.  There  are  approximately  68  primary  and  122  secondary  ribs  on  the  last 
whorl.  At  55  mm.  diameter  there  are  an  estimated  52  ribs;  at  60,  est.  54;  65,  est.  56; 
70,  est.  57  75,  est.  58  80,  est.  59.  85,  est.  60;  90,  est.  61;  95,  est.  63;  100,  est.  64  ; 
105,  est.  65;   no,  est.  66;   115,  est.  67.     (Text-fig.  6). 

The  ribs  on  the  inner  whorls  appear  to  be  of  a  style  similar  to  those  of  the  micro- 
conch ;  on  the  last  umbilical  whorl  they  appear  straight  and  rectiradiate  throughout 
their  length.  On  the  last  half  whorl  the  primary  ribs  become  more  widely  spaced 
and  prominent,  and  their  furcation  becomes  irregular.  Several  polygyrate,  simple 
and  intercalatory  ribs  are  developed. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  59 

The  peristome  is  simple. 

Remarks.  Since  the  inner  whorls  are  poorly  preserved,  the  macroconch  cannot 
be  matched  with  the  microconch  as  far  as  rib  densities  of  the  inner  whorls  go,  but 
their  association  and  similar  sculpture  strongly  suggest  that  they  are  dimorphs  of 
the  same  species. 

This  species  is  distinguished  from  others  by  its  rib  style  and  density.  It  bears  some 
resemblance  to  P.  (V.)  reisiformis  described  above  (p.  50)  in  rib  density,  but  the  ribs 
themselves  are  considerably  more  slender  in  P.  (V.)  encombensis .  The  macroconchs 
are  not  so  variocostate  as  those  of  P.  (V.)  reisiformis,  which  provides  another  basis 
for  distinction  between  these  two  species. 


Subgenus  PECTINATITES  Buckman  1922 

Synonyms.     Wheatleyites    Buckman    1923  ;     Keratinites    Buckman    1925  ;     ?    Pectiniformites 
Buckman  1925  (see  p.  20). 

Type  species.     (By  original  designation).     Ammonites  pectinatus  Phillips  1871. 

Diagnosis.  Dimorphic.  Microconchs  generally  finely  ribbed  on  inner  whorls, 
body-chamber  usually  more  coarsely  ribbed.  Peristome  with  ventral  horn  often  of 
great  length.  Macroconchs  generally  finely  ribbed  on  inner  whorls.  Outer  whorls 
variable,  primary  ribs  typically  strong  with  variable  number  of  secondary  ribs. 
Never  truly  virgatotome.  Both  macroconch  and  microconch  show  tendency  for 
ribs  to  bifurcate  very  low  on  whorl  side.     Constrictions  generally  absent. 

Upper  Kimmeridgian,  PWheatleyensis  Zone,  Pectinatus  Zone. 

Ammonites  pectinatus  Phillips  was  the  first  species,  now  included  in  this  genus, 
to  be  described.  Phillips'  figure  (1871,  pi.  15,  fig.  17)  is  very  poor,  and  the  holotype 
has  been  long  thought  to  have  been  lost.  Arkell  (1956  :  780)  therefore  designated  a 
topotype  as  the  neotype. 


Pectinatites  (Pectin atites)  inconsuetus  sp.  nov. 

(PL  30) 

Diagnosis.  Macroconchs  approximately  150  mm.  in  diameter.  Ribs  on  outer 
whorl  bifurcate  low  on  whorl  side.  Ornamentation  gradually  fading  on  body- 
chamber.  Peristome  straight.  Microconchs  approximately  100  mm.  in  diameter. 
At  50  mm.  diameter  there  are  approximately  50  ribs.  Point  of  bifurcation  of  ribs 
becoming  gradually  lower  on  whorl-side  towards  last  whorl.  Body-chamber  more 
coarsely  ribbed  than  inner  whorls.     Peristome  bearing  short  ventral  horn. 

Holotype.     Macroconch  C. 73446. 
Paratype  (allotype).     Microconch  C. 73447. 

Material.     Two  specimens,  both  plaster  casts  (one  macroconch,  one  microconch. 
Horizon.     Both  specimens  from  10  ft.  above  the  Middle  White  Stone  Band. 
Upper  Kimmeridgian,  lower  Pectinatus  Zone. 


60  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Description.  Macroconch.  E volute  shell  with  a  diameter  of  153  mm.  Diameter 
of  umbilicus  65  mm.  There  are  26  primary  and  approximately  83  secondary  ribs 
on  the  last  whorl. 

The  ribs  on  the  inner  whorls  are  not  well  preserved,  but  are  seen  to  be  rursiradiate 
at  the  umbilical  shoulder,  then  swinging  forwards  to  become  straight  and  slightly 
prorsiradiate.  The  primary  ribs  on  the  outer  whorl  at  first  become  more  pronounced 
and  are  more  widely  spaced.  They  branch  very  low  on  the  whorl-side,  giving  rise 
to  up  to  four  secondary  ribs.  There  are  abundant  intercalatory  secondary  ribs, 
which  also  arise  very  low  on  the  whorl-side. 

Over  the  last  quarter  of  a  whorl,  the  ribs  gradually  fade  and  become  very  indistinct. 

The  peristome  which  is  straight  and  simple  inclines  anteriorly  towards  the  venter. 

Microconch.  Evolute  shell  with  a  diameter  of  98  mm.  Diameter  of  the  umbilicus 
32  mm.  There  are  50  primary  and  an  estimated  no  secondary  ribs  on  the  last  whorl. 
There  are  approximately  50  ribs  at  a  diameter  of  30  mm. 

The  ribs  on  the  inner  whorls  are  fine  and  slender.  They  are  rursiradiate 
at  the  umbilical  shoulder,  then  curve  forwards  to  become  rectiradiate  or  slightly 
prorsiradiate.     The  point  of  bifurcation  of  the  ribs  is  high  on  the  whorl-side. 

The  umbilical  seam  uncoils  over  the  last  half  whorl. 

The  ribs  on  the  outer  whorl  are  similar  in  style  to  those  of  the  inner  whorls,  but 
gradually  the  point  of  bifurcation  of  the  ribs  becomes  much  lower  on  the  whorl-side. 
Over  the  last  half-whorl  the  primary  ribs  become  stronger  and  more  widely  spaced, 
and  usually  give  rise  to  three  secondary  ribs  on  the  whorl-side.  There  are  several 
simple  and  intercalatory  ribs  on  the  last  half  whorl. 

The  peristome  bears  a  horn.  On  the  paratype  this  is  broken,  and  the  resultant 
broken  end  has  not  reproduced  well  in  the  plaster.  The  basal  4  mm.  of  the  horn 
are  just  visible,  however. 

Remarks.  I  earlier  referred  the  macroconch  of  this  species  to  the  Tithonian  genus 
Pseudovirgatites  (Cope  &  Zeiss  1964  :  12).  At  the  time  of  making  this  identification, 
however,  the  microconch  had  not  been  discovered.  The  style  of  ribbing  of  the  macro- 
conch is  very  similar  to  some  specimens  of  Pseudovirgatites  from  Franconia.  How- 
ever, the  microconch,  with  its  broadly  similar  ribbing  on  its  body-chamber  and 
horned  peristome,  shows  that  this  species  belongs  to  the  genus  Pectinatites. 

This  again  is  an  example  of  the  remarkable  homeomorphy  between  the  Tithonian 
and  Upper  Kimmeridgian  ammonites  which  has  misled  so  many  workers  in  the  past. 
This  species  of  Pectinatites  with  its  type  of  modification  of  the  ribbing  on  the  body- 
chamber  of  the  macroconch  is  unlikely  to  be  confused  with  any  other  species. 

Pectinatites  (Pectinatites)  eastlecottensis  (Salfeld) 
(PI.  26,  fig.  1) 

1913  Perisphinctes  eastlecottensis  Salfeld  :  429,  pis.  41,  42. 

1914  Perisphinctes  eastlecottensis  Salfeld  ;    Salfeld  :  130. 

1922  Wheatleyites  eastlecottensis  (Salfeld)  Buckman  :  28. 

1923  Pectinatites  aulacophorus  Buckman,  pi.  381. 

1925     Wheatleyites  eastlecottensis  (Salfeld)  ;    Neaverson  :  37. 
x933     Pectinatites  eastlecottensis  (Salfeld)  Arkell :  457. 


UPPER    KIMMERIDGE   CLAY    OF    DORSET  61 

Material.     Two  specimens,  plaster  casts  (both  microconchs). 

Horizon.  Ten  feet  above  the  Middle  White  Stone  Band.  Upper  Kimmeridgian, 
lower  Pectinatus  Zone. 

Description.  Neither  of  the  two  specimens  from  Kimmeridge  is  complete. 
The  more  complete  of  the  two  (0.73449,  PI.  26,  fig.  1)  has  a  diameter  of  70  mm.  and 
an  umbilical  diameter  of  approximately  22  mm.  On  the  last  half  whorl  preserved 
there  are  an  estimated  58  primary  and  97  secondary  ribs.  This  would  mean  that  at  a 
diameter  of  70  mm.  there  are  about  100  ribs  on  a  complete  whorl. 

The  holotype,  which  is  a  macroconch,  has  130  ribs  at  no  mm.  diameter.  The 
holotype  of  P.  aulacophorus  Buckman  has  (according  to  Buckman)  about  97  ribs  at 
66  mm.  diameter. 

The  noticeable  feature  on  the  Dorset  specimens  is  that  the  ribs  often  bifurcate 
very  close  to  the  umbilical  shoulder. 

The  peristome  is  not  preserved  on  either  of  the  Dorset  specimens  but  was  probably 
originally  horned. 

Remarks.  The  extremely  dense  ribbing  of  this  species  is  very  characteristic. 
There  can  be  little  doubt  that  these  Dorset  specimens  are  the  microconch  of  Salfeld's 
figured  macroconch. 

The  holotype  was  quoted  by  Salfeld  as  coming  from  the  Upper  Lydite  Bed  at 
Swindon.  Chatwin  &  Pringle  (1921  :  166)  later  showed  that  in  fact  it  came  from 
the  upper  part  of  the  Shotover  Grit  Sands.  Buckman's  species  P.  aulacophorus  was 
quoted  by  him  as  occurring  in  his  Bed  12  at  Swindon — the  bed  which  yielded  the  holo- 
type of  P.  eastlecottensis. 

Buckman's  figure  shows  that  the  last  sutures  of  P.  aulacophorus  are  somewhat 
approximated,  and  that  the  umbilical  seam  is  just  beginning  to  uncoil.  In  this  case 
it  would  appear  that  only  the  body-chamber  is  missing  from  this  specimen,  and  it  is, 
therefore,  a  microconch.  Neaverson's  figure  (1925,  pi.  1,  fig.  5)  of  P.  aulacophorus 
is  an  immature  specimen  of  a  species  of  the  subgenus  Virgatosphinctoides  close  to 
P.  (V.)  wheatleyensis  delicatulus. 

The  occurrence  of  this  species  in  Dorset  enables  good  correlations  to  be  made  with 
Swindon  and  Oxford. 


Pectinatites  (Pectinatites)  cf.  groenlandicus  (Spath) 
(PI-  3i) 

1936     Pectinatites  [Keratinites?)  groenlandicus  Spath  :  25,  pi.  6,  fig.  1. 
Material.     One  specimen  (macroconch). 

Horizon.  Ten  feet  above  the  Middle  White  Stone  Band.  Upper  Kimmeridgian, 
lower  Pectinatus  Zone. 

Description.  The  single  incomplete  specimen  from  Kimmeridge  has  a  crushed 
diameter  of  380  mm.  Diameter  of  umbilicus  177  mm.  At  40  mm.  diameter  there 
are  48  ribs;  at  50,  53;  60,70,80,90,51;  100,52;  110.120,53;  130,54;  140,55; 
150,  54;    160,  53;    170,  52. 


62  UPPER   KIMMERIDGE   CLAY    OF   DORSET 

There  is  a  very  close  comparison  between  the  Dorset  specimen  and  the  holotype 
which  is  from  the  Pectinatus  Zone  of  Greenland.  The  furcation  and  style  of  the 
ribs  on  the  inner  whorls  is  very  similar  to  that  of  the  holotype  (Spath,  pi.  7,  fig.  5). 
The  point  of  bifurcation  is  very  high  on  the  whorl-side.  The  outer  whorl  is  similar, 
too,  with  the  ribs  becoming  less  prominent  towards  the  aperture  of  the  shell.  The 
peristome,  which  is  not  preserved,  is  presumably  simple. 

Remarks.  "  Wheatleyites  "  reductus  (Buckman)  (1923,  pi.  384)  shows  certain 
similarities  to  this  species,  but  the  inner  whorls  of  this  species  are  more  sharply  and 
densely  ribbed. 

Pectinatites  (Pectinatites)  cornutifer  (Buckman) 

(PI.  25,  fig.  3;   PL  26,  fig.  2) 

1925  Kevatinites  cornutifer  Buckman  :   pi.  602. 

1926  Kevatinites  nasutus  Buckman  :   pi.  664. 

Material.     Eight  specimens  (all  microconchs). 

Stratigraphical  range.  Upper  Kimmeridgian,  Pectinatus  Zone  (middle  part), 
from  19  ft.  below,  to  6  ft.  above  the  Freshwater  Steps  Stone  Band. 

Description.  Microconch.  Moderately  evolute  shell  with  a  diameter  of  68-90 
mm.  Diameter  of  the  umbilicus  22-30  mm.  There  is  good  agreement  with  Buck- 
man's  figures  in  all  respects.     The  horn  is  long  and  varies  from  15  to  39  mm.  in  length. 

No  macroconchs  have  been  found  at  this  horizon  in  Dorset  (see  below). 

Remarks.  K.  cornutifer  Buckman  is  finer  ribbed  than  K.  nasutus  Buckman. 
However,  there  is  a  complete  transition  in  the  Dorset  specimens  between  these  two 
forms.  The  cornutifer  type  occurs  in  the  lower  part,  and  the  nasutus  type  in  the 
upper  part  of  the  range  of  this  species  in  Dorset.  In  the  almost  complete  absence  of 
macroconchs  in  collections  from  this  horizon  in  Dorset,  it  is  not  possible  to  refer 
this  species  to  one  of  the  known  macroconch  species  of  Pectinatites.  P.  pectinatus 
(Phillips)  occurs  together  with  this  species  at  Swindon  and  in  the  Oxford  region,  so 
that  it  may  well  be  the  macroconch  of  P.  cornutifer. 

Pectinatites  (Pectinatites)  naso  (Buckman) 
(PL  32) 

1926     Keratinites  naso  Buckman,  pi.  652. 

Material.     Three  specimens,  all  plaster  casts,  (two  macroconchs,  one  microconch) . 

Horizon.  Ten  feet  above  the  Freshwater  Steps  Stone  Band.  Upper  Kimmerid- 
gian, middle  Pectinatus  Zone. 

Description.  Macroconch.  Fairly  evolute  shell  with  a  diameter  of  130-138  mm 
The  diameter  of  the  umbilicus  is  50-54  mm.  The  figured  specimen  (C. 73452)  which 
has  a  diameter  of  138  mm.  and  an  umbilical  diameter  of  54  mm.  has  40  primary 
ribs  on  the  last  whorl.  At  25  mm.  diameter  there  are  about  45  ribs;  at  30,  46;  35, 
47;   40,  49;  45,  47;   50,45- 


UPPER    KIMMERIDGE   CLAY    OF    DORSET  63 

The  ribs  on  the  inner  whorls  are  rectiradiate  at  the  umbilical  shoulder,  then  curve 
forwards  to  become  quite  strongly  prorsiradiate.  The  point  of  bifurcation  of  the 
ribs  is  high  on  the  whorl-side. 

On  the  outer  whorl,  just  over  half  of  which  is  preserved  on  the  figured  specimen, 
the  primary  ribs  become  stronger  and  more  widely  spaced.  They  branch  fairly  low 
on  the  whorl-side,  giving  rise  to  two  or  three  secondary  ribs.  There  are  very 
occasional  simple  and  intercalatory  ribs. 

The  peristome  is  simple. 

Microconch.  The  microconch  (C  .73453)  is  very  similar  to  those  figured  by  Buckman 
(pis.  652,  652a). 

The  Dorset  specimen  is  91  mm.  in  diameter,  and  has  an  umbilical  diameter  of  31 
mm.  There  are  an  estimated  47  primary  ribs  on  the  last  whorl.  The  preservation  of 
the  inner  whorls  is  not  good,  and  the  following  rib  densities  are  only  approximate : 
at  20  mm.  42  ribs;   25,  44;   30,  44. 

The  ribs  on  the  inner  whorls  are  similar  in  style  to  those  of  the  macroconch.  The 
umbilical  seam  uncoils  over  the  last  half  whorl.  On  the  outer  whorl  the  ribs  become 
rather  suddenly  more  coarse  half  a  whorl  from  the  aperture.  This  last  half  whorl 
appears  to  correspond  to  the  length  of  the  body-chamber  (to  judge  by  differences  in 
the  degree  of  crushing).  Buckman's  specimens,  too,  have  a  body-chamber  half  a 
whorl  in  length. 

On  the  last  half  whorl  the  primary  ribs  become  more  widely  spaced;  their  point 
of  furcation  is  lower  on  the  whorl-side  than  on  the  inner  whorls,  and  the  angle  of 
furcation  increases.  Most  of  the  ribs  on  the  body-chamber  are  bifurcate,  but  there 
are  occasional  unbranched  primary  ribs. 

The  peristome  bears  a  horn  which  is  21  mm.  long. 

Remarks.  This  species  is  readily  distinguishable  from  P.  cornutifer  described 
above  by  the  more  coarsely  ribbed  body-chamber  of  the  microconch. 

The  macroconch  is  considerably  more  coarsely  ribbed  than  the  macroconch  of 
P.  pectinatus. 

Subfamily  DORSOPLANITINAE  Arkell  1950 

Genus  PAVLOVIA  Ilovaisky  1917 
Subgenus  PARAVIRGATITES  Buckman  1922 
Pavlovia  (Paravirgatites)  cf.  paravirgatus  (Buckman) 

(PI-  33) 
1922     Paravirgatites  paravirgatus  Buckman,  pi.  353. 

Material.     One  specimen  C. 73454. 

Horizon.  Ten  feet  above  the  Freshwater  Steps  Stone  Band.  Upper  Kim- 
meridgian,  Pectinatus  Zone  (middle  part). 

Description.  The  single  poorly  preserved  specimen  has  a  diameter  of  146  mm. 
and  an  umbilical  diameter  of  66  mm.  There  are  28  primary  and  an  estimated  55 
secondary  ribs  on  the  last  whorl. 


64  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Remarks.  The  specimen  agrees  closely  with  Buckman's  figure  in  all  respects 
except  size.  The  holotype  is  about  220  mm.  diameter.  The  general  similarity  and 
the  similar  horizon  (Shotover  Grit  Sands)  leave  little  doubt  of  the  affinities  of  the 
Dorset  specimen  to  Buckman's  holotype. 


V.     EVOLUTION    OF    THE    AMMONITES 

The  possible  origin  of  the  genus  Pectinatites  from  an  ataxioceratid  stock  has  been 
discussed  earlier  (p.  22).  Within  the  genus  three  subgenera  are  recognized  and  it  is 
possible  to  follow  in  these  subgenera  various  evolutionary  trends. 

The  subgenus  Arkellites  first  appears  at  the  base  of  the  Elegans  Zone  where  it  is 
represented  by  specimens  referable  to  P.  (A.)  primitivus.  Arkellites  is  characterized 
by  more  or  less  equicostate  ribbing  of  the  shell.  This  feature  appears  in  the  four 
species  of  the  subgenus  hitherto  described.  The  macroconchs  remain  basically 
similar  in  rib  style  and  ornamentation  throughout  the  succession.  In  the  micro- 
conchs,  however,  the  tendency  is  for  the  horn  to  become  much  more  prominent.  P. 
{A.)  primitivus  has  a  weak  ventral  peristomal  inflation,  but  all  the  later  species  have 
a  well-developed  true  horn.  The  youngest  species  of  Arkellites  hitherto  recorded, 
P.  {A.)  hudlestoni,  shows  a  general  similarity  to  these  earlier  species,  and  the  conclu- 
sion is  drawn  that  this  subgenus  was  a  fairly  conservative  one.  There  are  as  yet  no 
species  recorded  from  the  Wheatleyensis  Zone  which  can  with  certainty  be  placed 
in  this  subgenus,  so  that  the  connection  between  P.  (A.)  hudlestoni  and  earlier  species 
of  the  subgenus  is  not  known.  However,  Paravirgatites  kimmeridgensis  Neaverson 
(1925  :  33,  pi.  4,  fig.  4)  has  a  rib  density  on  its  inner  whorls  close  to  that  of  P.  {A.) 
hudlestoni,  and  may  belong  to  this  subgenus.  The  author  cannot  accept  Neaverson's 
placing  of  his  species  in  Buckman's  genus  Paravirgatites.  The  holotype  shows  the 
rib  style,  rib  density  and  development  typical  of  Arkellites,  which  is  entirely  different 
from  the  sharp  regular  bifurcate  ribbing  characteristic  of  the  pavlovids.  It  is  to  be 
expected  that  the  Wheatleyensis  Zone  will  ultimately  yield  species  which  can 
definitely  be  assigned  to  the  subgenus  Arkellites. 

The  subgenus  Virgatosphinctoides  which  appears  in  Dorset  about  a  third  of  the  way 
up  the  Elegans  Zone  could  have  been  derived  from  P.  (A.)  primitivus  (see  p.  33). 
Unlike  Arkellites,  from  which  it  probably  arose,  Virgatosphinctoides  evolved  rapidly. 
The  horn  of  the  microconch,  represented  by  a  ventral  peristomal  inflation  in  P.  (V.) 
elegans,  becomes  a  true  horn  by  the  top  of  the  Elegans  Zone  (P.  (V.)  elegans  corniger). 
Thereafter,  the  horn  development  becomes  more  pronounced,  particularly  in  the 
Hudlestoni  Zone.  There  are  some  species,  however,  (e.g.  P.  (V.)  woodwardi)  in 
which  the  horn  development  is  not  so  pronounced.  Considering  next  the  macro- 
conchs, the  tendency  seen  is  for  the  degree  of  variocostation  of  the  shell  to  become 
more  pronounced.  Associated  with  this  is  the  increase  in  the  numbers  of  polygyrate 
ribs  on  the  body-chamber.  This  trend  continues  with  the  appearance  of  virgatotome 
ribbing  in  the  Wheatleyensis  Zone.  Some  later  forms  from  the  Hudlestoni  Zone 
(e.g.  P.  (V .)  donovani)  show,  to  some  extent,  a  reversal  of  this  trend,  and  the  loss  of 
the  truly  virgatotome  rib  type. 


UPPER  KIMMERIDGE  CLAY  OF  DORSET 


65 


Fig.  8.     Possible  phylogenetic  relationships  of  species  of  Pectinatites . 


geol.  15,  1. 


66  UPPER    KIMMERIDGE   CLAY    OF   DORSET 

The  origins  of  the  subgenus  Pectinatites  are  almost  certainly  to  be  found  in  the 
subgenus  Virgatosphinctoides.  However,  the  actual  point  of  origin  is  not  clear. 
Pectinatites  may  be  derived  from  such  a  form  as  P.  (V.)  encombensis ,  which  has  a 
finely  ribbed  shell,  a  microconch  with  a  well-developed  horn,  and  an  absence  of 
virgatotome  ribs  on  the  macroconch.  Alternatively,  the  subgenus  may  have  been 
derived  from  Virgatosphinctoides  earlier,  perhaps  as  early  as  the  Wheatleyensis  Zone. 
In  Pectinatites  the  main  trends  observed,  as  far  as  the  lower  part  of  the  Pectinatus 
Zone  goes,  appear  to  be  the  tendency  for  the  ribs  to  bifurcate  low  on  the  whorl-side, 
and  for  the  microconch  horn  to  increase  in  length,  reaching  almost  40  mm.  in  P.  {P.) 
comutifer. 

Homeomorphy  occurs  within  the  genus  Pectinatites,  particularly  between  species 
of  the  subgenera  Virgatosphinctoides  and  Pectinatites.  However,  the  homeomorphy 
always  seems  to  appear  in  only  one  sex,  and  not  in  both  sexes  together. 

Text-fig.  8  shows  diagrammatically  the  possible  relationships  between  the  known 
species  of  the  genus.  It  is  not,  of  course,  to  be  expected  that  this  can  represent  an 
entirely  complete  picture.  In  particular,  knowledge  is  scant  at  the  base  of  the 
Elegans  Zone,  the  upper  part  of  the  Scitulus  Zone,  and  the  lowermost  Pectinatus 
Zone. 

VI.     THE    AMMONITE    ZONES 

As  a  result  of  detailed  collecting  from  the  Kimmeridge  section,  the  position  of 
several  species  of  ammonites,  described  from  other  areas,  has  been  established  for 
the  first  time.  This  has  necessitated  considerable  modification  of  the  existing  table 
of  zones  (see  Text-fig.  9). 

Since  the  range  of  many  species  is  known  fairly  accurately,  it  is  proposed  to  set  up 
a  sub-zonal  scheme,  should  this  prove  possible,  at  a  future  date.  As  a  prerequisite 
for  this,  however,  detailed  knowledge  of  the  ammonite  faunas  over  a  large  area  is 
considered  necessary.  Unfortunately  the  Oxford  and  Swindon  areas,  which  could 
have  yielded  much  from  careful  collecting,  are  now  devoid  of  good,  or  even  adequate, 
exposures  of  Upper  Kimmeridge  Clay.  The  sections  in  Yorkshire  and  Sutherland 
may,  however,  provide  good  information,  particularly  on  the  lower  zones. 

At  this  time,  therefore,  no  further  refinement  than  zonal  subdivision  is  attempted. 

Pectinatites  ( Virgatosphinctoides)  elegans  Zone 

Index  species.     Pectinatites  (Virgatosphinctoides)  elegans. 

This  new  zone  is  proposed  for  the  beds  between  the  thin  cementstone  band  (Bed 
no.  42)  and  the  Yellow  Ledge  Stone  Band  (Bed  no.  36)  of  the  Kimmeridge  section. 

This  new  zonal  index  replaces  a  zone  based  on  species  of  the  genus  Gravesia  which 
were  first  used  for  zonal  subdivision  of  the  Kimmeridge  Clay  by  Salfeld.  After  his 
discovery  in  Dorset  of  species  of  ammonites,  for  which  he  proposed  the  genus,  he 
set  up  two  zones,  for  the  shales  between  the  Maple  Ledge  Stone  Band  (Arkell  1947  :  73) 
and  the  Yellow  Ledge  Stone  Band,  with  species  of  Gravesia  as  their  index  fossils.  He 
proposed  an  upper  zone  of  Gravesia  irius  and  a  lower  one  of  Gravesia  gravesiana 
(Salfeld  1913). 


UPPER  KIMMERIDGE  CLAY  OF  DORSET 


67 


cc 

< 


Lu 

to 
n 

1— . 

K. 

T 

H 

P- 

-» 

h~ 

t> 

O 

III 

Ct 

0, 

Uj 

•=1 

0. 

CO 

10 


01 
Uj 

CL 

s 

10 


n 

to 

>n 

Uj 
i~. 

5 

5 

S 

S 

i. 

-1 

u 

in 

0 

10 

Q. 

n 

(/> 

01 

Ct 


i75 


CO 

Uj 
to 


z 
O 
00 

CC 

Ul 

> 
< 

UJ 


Uj 


to 

>-. 
I 
CO 
ft 


10 

Ul 

to. 

10 

0 

t— 

to 

C) 

<• 

> 

Ul 

>- 

J- 

Ul 

u 

-J 

tO 

0 

s 

I— 

Ul 

■<r 

-c 

0 

•< 

ct 

01 
UJ 

z 
O 

tsl 


to 
uj 
:*. 
<* 
ct 

ID 


in 
u) 

z: 
O 
N 


LO 

< 

I 
CL 
UJ 

h- 
LO 

o 

U 
< 

_i 

< 


CO 

Uj 

to 

K- 

0 

0) 

O 

^ 

T— 

2 

1 

0 

a: 

to 

-J 

CL 

<* 

UJ 

to 

-j 

L_ 

-j 

_1 

f* 

1 

< 

01 

Uj 
CL 

u. 

to 

Uj 


to 

Uj 

1 

to 
5 


oi 
3 


g 
Uj 

ct 

CD 


oi 
to 


to 

UJ 

z 
O 

M 

Q 
Ul 

to 

O 
CL 
O 
CC 
0_ 


oi 

I— 


0 

ft 


Uj 

Ct 


2; 
o 

Uj 
-~J 

Q 

5 


g 

Co 

Uj 
CL 


to 
Uj 
Q 


Co 

2: 

O 

CO 

Ct 


to 

o 

I— 

"=t 

CD 

ct 


o 
to 


<: 

CL 

to 
o 

(^ 

"=? 

CD 

Ct 


Fig.  9.     Zones  of  the  lower  part  of  the  Upper  Kimmeridge  Clay, 


68  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

Salfeld  did  not  draw  any  junction  between  these  two  zones,  and  the  specific 
identity  of  his  specimens  has  long  been  in  doubt.  Arkell  (1947  :  76)  reported  that 
he  had  seen  Salfeld's  specimens  in  Gottingen  in  1937,  but  did  not  comment  upon 
their  identity.  He  called  these  two  zones  merely  the  Gravesia  spp.  zones  (1947  :  67), 
and  raised  the  lower  limit  of  the  zones  in  Dorset  up  to  the  unnamed  cementstone  band 
at  the  foot  of  Hen  Cliff  (about  65  ft.  higher  in  the  succession). 

Later  Arkell  (1956  :  21)  divided  these  Hen  Cliff  shales  into  an  upper  Gravesia  gigas 
Zone  and  a  lower  Gravesia  gravesiana  Zone.  Again,  no  boundary  between  these 
zones  was  fixed. 

Arkell's  raising  of  the  upper  limit  of  the  Aulacostephanus  Zones  is  justified  by  the 
occurrence  of  this  genus  up  to  about  15  ft.  below  the  thin  cementstone  band  referred 
to  above.  This  band  also  marks  the  first  appearance  of  specimens  of  Pectinatites 
(Arkellites),  which  are  fragmentary  and  poorly  preserved,  but  are  probably  close  to 
P.  (A.)  primitivus. 

The  genus  Gravesia  is  exceptionally  rare  in  Dorset,  and  since  1913  only  seven  speci- 
mens of  the  genus  have  been  found  (three  by  Spath,  two  of  which  are  in  the  British 
Museum,  and  one  in  the  Geological  Survey  Museum;  and  four  by  the  author). 
Of  these  ammonites,  five  are  specimens  of  Gravesia  gigas  and  have  been  found  between 
40  and  52  ft.  below  the  Yellow  Ledge  Stone  Band.  The  other  two  are  referred  to 
Gravesia  gravesiana  and  came  from  eight  feet  below  and  six  feet  above  the  Yellow 
Ledge  Stone  Band  (i.e.  higher  than  any  previously  recorded  specimens).  No  known 
Gravesia  irius  has  been  found  since  Salfeld's  report  of  its  abundant  occurrence  in  1913. 

Whether  or  not  Gravesia  irius  does  in  fact  occur  at  Kimmeridge,  it  is  clear  that 
Gravesia  gravesiana  is  restricted  to  beds  higher  than  those  yielding  Gravesia  gigas 

It  is  thus  proposed  to  set  up  this  zone  based  on  a  species  of  Pectinatites,  since 
species  of  this  genus  are  common  in  these  beds  in  Dorset.  This  obviates  any  diffi- 
culty over  fixing  of  boundaries  of  zones  based  on  extremely  rare  index  fossils,  and 
since  species  of  Pectinatites  occur  in  this  zone  in  Yorkshire  (whereas  Gravesia  does  not) 
there  should  in  future  be  no  ambiguity  as  in  the  past. 

The  base  of  the  zone  is  fixed  above  the  highest  occurrence  of  Aulacostephanus, 
and  at  the  earliest  occurrence  of  Pectinatites.  P.  {Arkellites)  primitivus  occurs  from 
the  base  of  the  zone  into  the  upper  part.  P.  (V.)  elegans  appears  below  the  middle  of 
the  zone  and  ranges  into  the  upper  part,  where  it  is  replaced  by  P.  (V.)  elegans 
corniger.  P.  (Arkellites)  cuddlensis  occurs  in  the  top  18  ft.  of  the  zone.  The  top  of 
the  zone  corresponds  to  the  highest  occurrence  of  P.  (V.)  elegans  corniger.  Gravesia 
gigas  occurs  just  below  the  middle  of  the  zone,  G.  gravesiana  ranges  from  the  highest 
part  of  the  zone  into  the  base  of  the  succeeding  Scitulus  Zone. 

Pectinatites  (Virgatosphinctoid.es)  scitulus  Zone 

Index  species.     Pectinatites  (Virgatosphinctoid.es)  scitulus  sp.  nov. 

This  new  zone  is  proposed  for  the  shales  between  the  Yellow  Ledge  Stone  Band 
and  the  Grey  Ledge  Stone  Band  in  the  Dorset  succession.  This  thickness  of  90  ft. 
includes  the  Lower  Cattle  Ledge  Shales  (up  to  Cattle  Ledge),  and  the  Upper  Cattle 
Ledge  Shales  (between  Cattle  Ledge  and  Grey  Ledge). 


PECTINATUS 


Pectinatites  (  Pectinatites)   inconsuetus 

H 


P.  (P.)  naso 
H 


P(P)  cornutifer 


P.  (  P)  groenland  icus 
H 


P(R)  eastlecottensis 
1 


Pavlovia  (Paravirgati  tes)  paravirgatus 


PECTINATUS 


■q|j^iiUllli!illi!lllll!lll^ 

-H-CD 


(Paravtrgatites)  paravirgatu: 


Fig.  io.    Stratigraphical  ranges  ol  the  ammonites  from  the  lower  part  of  the  Upper  Kimmeridge  Clay, 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  69 

The  Upper  Cattle  Ledge  Shales  have  hitherto  failed  to  yield  ammonites  and  are 
provisionally  included  in  this  zone,  pending  ammonite  evidence  from  Dorset  or 
elsewhere. 

Pectinatites  (V  irgatosphinctoides)  scitalus  first  appears  at  the  base  of  the  zone,  and 
ranges  to  about  the  middle  of  the  ammonite  bearing  strata.  Gravesia  gravesiana 
occurs  in  the  lowest  six  feet  of  the  zone  at  Kimmeridge.  Exogyra  virgula  ranges  up 
to  27  ft.  above  the  Yellow  Ledge  Stone  Band.  Lingula  ovalis  reaches  its  maximum 
abundance  near  the  base  of  the  zone. 

This  zone  corresponds  to  the  lower  part  of  Salfeld's  Virgatites  miatschkovensis 
Zone,  the  lower  part  of  Neaverson's  V irgatosphinctoides  wheatleyensis  Zone,  and  the 
Subplanites  ?  vimineus  Zone  of  Spath  (Arkell  1956  :  21). 

Pectinatites  (Virgatosphinctoides)  wheatleyensis  Zone 

Index  species.     Pectinatites  (Virgatosphinctoides)  wheatleyensis  (Neaverson). 

This  zone,  which  in  the  Dorset  succession  is  represented  by  the  beds  between  the 
Grey  Ledge  Stone  Band  and  the  Rope  Lake  Head  Stone  Band,  corresponds  to  part 
of  the  Virgatites  miatschkovensis  Zone  of  Salfeld  (1913) ;  the  Pseudovirgatites  Zone,  of 
Lamplugh,  Kitchin  &  Pringle  (1922) ;  the  lower  part  of  the  Pectinatus,  the  Nodiferus 
and  the  upper  part  of  the  Wheatleyensis  Zone  of  Neaverson  (1925) ;  and  to  all  but 
the  uppermost  part  of  the  Grandis  Zone  of  Arkell  (1956). 

The  position  of  Pectinatites  (Virgatosphinctoides)  wheatleyensis  (Neaverson)  has 
hitherto  been  very  uncertain  in  the  Dorset  succession.  In  Oxfordshire,  it  occurs 
associated  with  P.  (V.)  woodwardi,  P.  (V.)  wheatleyensis  delicatnlus  and  Sphinctoceras. 
The  same  faunal  association  (without  Sphinctoceras) ,  but  with  P.  (V.)  grandis  and 
P.  (V.)  pseudoscruposus  in  addition,  is  found  in  Dorset  in  the  shales  immediately 
below  the  Blackstone. 

At  the  base  of  the  zone  ammonites  of  the  subgenus  Virgatosphinctoides  are  repre- 
sented by  P.  (V.)  clavelli,  P.  (V.)  smedmorensis  and  P.  (V.)  laticostatus;  the  latter 
two  species  ranging  up  to  the  middle  of  the  zone. 

The  crinoid  Saccocoma  ranges  in  Dorset  through  13  ft.  of  Beds  in  the  upper  part  of 
the  Zone. 

The  top  of  the  Zone  corresponds  to  the  highest  occurrence  of  P.  (V.)  wheatleyensis 
delicatulus  and  the  earliest  occurrence  of  P.  (V.)  reisiformis. 

The  zone  is  represented  in  the  Oxford  district  by  the  Wheatley  Nodule  Bed,  and  is 
present  on  the  Yorkshire  coast. 

Pectinatites  (Arkellites)  hudlestoni  Zone 

Index  species:  Pectinatites  (Arkellites)  hudlestoni  sp.  nov. 

This  new  zone  is  proposed  for  the  beds  between  the  Rope  Lake  Head  Stone  Band 
and  the  White  Band  in  the  Kimmeridge  succession.  It  corresponds  to  the  upper 
part  of  Salfeld's  Virgatites  miatschkovensis  Zone,  part  of  Neaverson's  Pectinatus 
Zone,  and  to  the  Wheatleyensis  Zone  and  topmost  part  of  the  Grandis  Zone  of 
Arkell.  At  the  base  of  the  zone  Pectinatites  (Virgatosphinctoides)  reisiformis  occurs, 
and  a  little  above  the  base  is  associated  with  P.  (A.)  hudlestoni  which  ranges  through- 


70  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

out  the  zone.  The  middle  part  of  the  zone  is  characterized  by  P.  [V '.)  donovani 
and  the  upper  part  by  P.  [A.)  hudlesioni,  P.  (V.)  encombensis  and  P.  [V .)  magnimas- 
citlus. 

Inland  this  zone  may  be  represented  by  the  Shotover  Fine  Sands  and  the  Lower 
Cemetery  Beds  in  the  Oxford  and  Swindon  areas  respectively.  There  is  no  palaeonto- 
logical  evidence  to  support  this  correlation  directly,  however,  and  the  zone,  if  present 
is  certainly  very  much  attenuated.     It  is  probably  present  in  Yorkshire. 

Pectinatites  (Pectinatites)  pectinatus  Zone 

Index  species.     Pectinatites  [Pectinatites)  pectinatus  (Phillips). 

This  zone  corresponds  to  the  Perisphinctes  pallasianus  Zone  of  Salfeld,  the  upper 
part  of  the  Pectinatus  Zone  of  Neaverson,  and  the  Pectinatus  Zone  of  Arkell. 

In  Dorset  the  lower  boundary  of  the  zone  is  taken  at  the  White  Stone  Band  which 
marks  the  upper  limit  of  the  range  of  the  subgenus  Virgatosphinctoides ,  and  the  upper 
boundary  below  the  first  occurrence  of  Pavlovia  s.s.  The  upper  part  of  this  zone  has 
not  yet  been  fully  investigated  in  Dorset. 

The  earliest  species  recorded  in  Dorset  is  Pectinatites  [Pectinatites)  eastlecottensis; 
this  species  is  recorded  together  with  such  species  as  P.  [P.)  cornutifer  and  P.  [P.)  naso 
from  the  Shotover  Grit  Sands  in  the  Oxford  region.  No  detailed  stratigraphical 
collections  have  been  made  from  these  beds  in  the  Oxford  region,  however,  and  it 
may  be  that  they  are  not  in  fact  completely  synchronous,  as  published  faunal  lists 
suggest.  In  Dorset  there  is  little  overlap  of  the  ranges  of  these  species  (Text-fig.  10), 
and  it  is  to  be  expected  that  detailed  collecting  would  show  similar  relationships 
between  the  various  ranges  of  species  in  the  Oxford  area,  where  the  succession  is 
considerably  thinner. 

VIII.     CORRELATIONS 

(a)  Great  Britain 

The  Upper  Kimmeridge  Clay  is  exposed  in  only  a  limited  number  of  localities  in 
Britain,  and  of  these  few  exposures  most  are  now  very  poor.  The  majority  of  the 
published  faunal  lists  from  these  exposures  are  now  outdated,  and  nowhere  have 
collections  been  made  in  detail  comparable  to  that  recently  carried  out  in  Dorset. 
For  these  reasons,  correlations  with  other  areas  of  Britain  cannot,  in  most  cases,  be 
established  with  a  great  deal  of  accuracy  at  present.  It  is  hoped  that  future  collecting 
will  remedy  this  deficiency. 

Two  areas  of  Britain  where  the  succession  of  the  Kimmeridgian  faunas  have  been 
well  known  for  some  time  are  the  Swindon  and  Oxford  regions.  Correlations  with 
these  areas  are  shown  in  Text-fig.  n.  In  both  these  areas  the  succession  is  consider- 
ably attenuated.  The  ammonites,  however,  are  generally  better  preserved  than 
those  in  Dorset  and  have  therefore  attracted  considerably  more  attention  in  the  past. 
It  was  primarily  on  information  obtained  from  the  Oxford  area  that  Neaverson  (1925) 
set  up  his  zonal  scheme  for  the  Upper  Kimmeridge  Clay.  It  is  now  possible  for  the 
first  time  to  show  the  true  stratigraphical  position  of  many  of  Neaverson's  species  in 
the  complete  Dorset  succession.     As  a  direct  result  of  this,  it  appears  likely  that  there 


UPPER  KIMMERIDGE  CLAY  OF  DORSET 


71 


)oo 

)lu 

( _i 

oo 

H 

00 

10           Q 

i-  I 

UJ                      ) 

^             J 

Uj                 ; 

0            ( 

UJ                     ) 

t    0 

*      i 

>     ^ 
0     °° 

1- 

UJ 
UJ 

cc 

UJ 

Q 

z 

1— / 

5       0 

uj        (  Z 

Ul   ( 
UJ   ( 

0 
cc 

0 

U_ 
X 
O 

u. 
0 

> 

0 

1- 

< 

00 

Ul) 
LU<i 

u_/ 

^        ( 
O        )  I 
UJ            1- 

LL    ) 
If)    ( 

Ct 

0 

~i 

5  5 

00 

0     *- 

CM 

0 

UJ 

i  r 
0  ( 00 

^ 

OO                         < 

Ul     -S 

^      cc 

00      0 

m 

I 
00 

z 
u_ 

\r ) 

0   ( 

5: 

0            ( 

-1     Cc 

<     >■ 
I     ^ 
,0     O 

O" 

5  1  >- 

2: 

< 

OO     JJ 

\  _i 

Uj 

( u 

W 

1 

cc 

1 

00    5 

00 

00    ) 

} 

f  y 

00 

Q 

a 

Q    ) 

UJ        ( 

Q 

"0 

Z       00 

z 

ui    \ 

0     (    00 

CC 

<     H 

N      — 
N      Q_ 
< 

CO     LlI 

z 
0 

Q 

<     Q 

00        Ul 
CD 
1- 

£     cc 

i- 

Ul 

Ul 

u. 

< 

00 

Ul 

z 

cD    1 

>-     ( 
cc     / 
u     ) 
1-     \ 

lb 

Ul      , 

W    (       U. 

Q 
CC 

< 

>- 

UJ                     / 

s        ( 

Lu             ) 

3            / 
O            ) 

CC 
< 

M 
M 
< 
CO 

1      UJ      ( 

S    ^    1 

0-)    3     ) 

(  0   \ 

Z 

O       UJ 

If) 

Li- 

Ul       ( 

i(       O 

^ 

Lu           { 

cc  (  UJ    ( 

_j 

£ 

rv      f- 

<tf 

CC 

2     / 

0 

00          / 

LL 

Ul   )    LO    / 

\k    ° 

^        UJ 

1 

UJ 

> 

0 

Ul        ) 

0  /       00 

0 

a.  )  >   ) 

O    a 

00 

>  2 

0    w 

If) 

U      \ 

^/       >- 

cc 

2      / 

0-  (Z.) 

en 

CC          ) 

CN.(             < 

CO 

5     ) 

00 

3     If 

00    I 
>- 

1-    0 

H 

Ul          (, 

1                U 

o-.   ) 

< 

\  C./' 

< 

O      cc 

O 

5      1 

_J 

_i 

I        Ul 

X 

0 

( 

0 

0 

00     a 

00 

-J 

*t  ( 

CL 

(^ 

— ' 

z> 

Ul 

wz° 

1-  CO    < 

00 

Q_    h- 
UJ    UJ 
1-    Ul 
00    Li- 

LU 

z 
Ul  0 

X  1- 

1-    00 

UJ 

1- 
X 

Ul    Ul    UJ 
X    Z    0- 

1-00 

LU 

z 

0  H 

00 
uj 
_i 
< 

00 

UJ 

00 

LU 
_l 
< 

CC   0 

5 

a 

t-  cc 

1-    UJ 

< 

UJ 

>EDS     ABOVE 
HITE    STONE 
rO  65      FEET 

ui  co 

Z   Q 

Z    H 

Z    00 

00  UJ 

X 

1- 

I 

H 

I 
LO 

0 

1—     v- 

Ul   < 

Ul 

<   Ul 

LLl           UJ 

u. 

LO 

Ul 

00 

Ul 

0 

E 
UJ 

H 

UJ  co 
cc 

LlI   uj 
£  X 

h- 

Ul  Ld 
CD   v£ 

< 

00   _j 

X 
I— 

Q 

z 
< 

CO   Ul 
Ll 

Ul 

O    CO 

H    r- 

00 

S    BETWE 
Y     LEDGE 
}   AND   TH 

0 
<  n 

UJ  00 
1 
a 

Ul   z 

*  < 

Ul 

0 
0 

UJ 

_l 

UJ 

Ul 
U. 

O 
C5) 

u_ 

LL 

Zi 
u 

z 

UJ 

LL 

n 

Ul 
O 
O 

Ul 

_l 

Ul 

z 

a 

a 

Q    liJ   " 
UJ    CC    ^ 

m  0  5 

<    CO 

_J 

Ul 

_l 

QJ§ 

■*     CL 

Ul  0 

Ul   Ul 

z 

_l 

h- 

I 

Q- 

I  1- 

CO  CL 

< 

h- 

< 

3 

1-   00 

0 

rO 

< 

2 

cc 

0 

-N 

00 

j 

LO 

UJ 

LO 

UJ 

UJ 

10 

UJ 

f        0^ 

00 

is 

(J 

01 

UJ 
1— 

to 

UJ 

I— 

0 

TINATITES 
rOSPHINCTOIC 

LO 
00 

UJ 

LO 
UJ 

Q 

6 

00 

Uj 

Q 

O 

5     2: 

00 

3:    00 
Q.    ui 

z 
0 

N 

l           5 

>-     c 

1 

Ul 

is 

01 
UJ 

Q 

Uj 

1 
S 

C 

5  2 

0      0 

ct     ° 

LO         UJ 

0     -j 

0^    O 
0    ^J 

0      0 

uj        Uj 

a.     a. 

UJ 

a. 

UJ 

3 
3; 

UJ 

t-~         LO 

cc 

Cj 

UJ 

fc!        Uj 

cc 

0 

—j 

S 

■t 

^. 

s 

Fig.   11.     Correlation  of  the  lower  part  of  the  Upper  Kimmeridge  Clay  of  Kimmeridge 
with  that  of  the  Swindon  and  Oxford  areas. 


72  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

is  a  previously  undetected  non-sequence  in  both  the  Oxford  and  Swindon  areas, 
between  the  Pectinatus  and  Wheatleyensis  Zones,  corresponding  to  the  Hudlestoni 
Zone.     This  zone  may  be  represented  in  part,  however,  by  the  Shotover  Fine  Sands. 

Between  the  Lower  and  Upper  Kimmeridge  Clay  there  is  a  major  non-sequence 
in  the  Oxford  area.  Above  the  Aulacostephanus  Zones  are  beds  yielding  Pectinatites 
(Virgatosphinctoides)  wheatleyensis.  There  are  no  known  records  of  fossils  indicative 
of  the  Elegans  or  Scitulus  Zones.  At  Swindon,  however,  the  position  is  different. 
Chatwin  &  Pringle  (1922  :  165)  mention  that  the  Hudleston  collection  contains  a 
specimen  of  Gravesia  from  Swindon,  and  it  is  therefore  possible  that  the  succession  in 
this  region,  though  attenuated,  is  fairly  complete. 

In  Yorkshire  the  Kimmeridge  Clay  is  exposed  beneath  the  Lower  Cretaceous  rocks 
which  rest  unconformably  on  it  at  Speeton.  The  Kimmeridge  Clay  here  is  little 
known  palaeontologically.  The  highest  horizon  recorded  in  this  section  is  the 
Wheatleyensis  Zone,  but  a  specimen  sent  to  me  by  Dr.  P.  Kaye  from  the  highest  beds 
of  the  Kimmeridge  Clay,  is  undoubtedly  close  to  Pectinatites  (P.)  proboscide  (Buck- 
man),  requiring  correlation  with  the  Pectinatus  Zone.  There  appears  no  reason  to 
believe  that  the  Yorkshire  succession  is  not  complete  up  to  this  latter  zone.  The 
Elegans  Zone  in  Yorkshire  is  unlikely  to  yield  Gravesia  since  the  presence  of  this 
genus  has  not  been  confirmed  north  of  Swindon. 

In  Sutherland,  the  lowest  part  of  the  Upper  Kimmeridge  Clay  appears  to  be  present, 
to  judge  from  the  account  by  Bailey  &  Weir  (1932).  Above  the  Aulacostephanus 
zones  they  recorded  species  of  Lithacoceras  indicative  of  the  Gravesia  zones,  and 
evidence  for  the  lower  part  of  the  Virgatites  zone.  I  interpret  this  evidence  as 
showing  that  the  Elegans  and  possibly  Scitulus  Zones  are  present.  The  record  of 
Lithacoceras  presumably  refers  to  a  species  of  Pectinatites  perhaps  of  the  Elegans  Zone. 
This  area  is  one  from  which  collecting  is  planned  in  the  future. 


(b)  The  Boulonnais 

The  Upper  Kimmeridgian  of  the  Boulonnais  shows  similarities  to  the  Kimmeridgian 
of  Britain.  It  is  remarkable  chiefly  for  the  development  of  phosphatic  nodule  beds 
at  several  horizons.  The  succession  below  is  based  on  descriptions  by  Pruvost  (1924), 
with  modifications  after  Arkell  (1956  :  42)  and  revised  determinations  of  the 
ammonite  names. 

Bed.  No.     (Pruvost)  1924) 

Ph.  3.     Tour  Croi  Nodule  Bed  with  phosphatized  ammonites: 

Pavlovia  rotunda,  P.  leblondi,  Pectinatites  (Pectinatites)  devillei,  P.   (P.) 

boidini,  Pectinatites  (P.)  rarescens,  P.  (P.)  opulentus. 
5.  Clay  26  ft.:   P.  (P.)  devillei,  P.  (P.)  boidini,  P.  (P.)  sp.,  Pavlovia  lydianites, 

Exogyra  dubiensis. 
Ph.  2.     Phosphatic  nodule  Beds:   undescribed  ammonites: 
4.  Clay  6  ft.  6  ins. :    undescribed  ammonites,  Lingtda  ovalis,  Discina  latissima, 

Modiola  autissiodorensis,  Anomia  laevigata. 


UPPER   KIMMERIDGE   CLAY   OF   DORSET  73 

Ph.  i.     La  Rochette  Nodule  bed:  Pectinatites  (Virgatosphinctoides)  pringlei,  P.  (V.) 

spp. 
3.  Clays  26  ft.  Discina  latissima,  Pectinatites  (Virgatosphinctoides)  spp. 

2.  Gres  de  la  Creche  (upper  part)  16  ft.:   Pectinatites  (Virgatosphinctoides)  sp. 

1.  Gres  de  la  Creche  (lower  part)  32  ft.  6  ins.     Gravesia  portlandica  (?  —G.  gigas), 

"  Perisphinctes  "  bleicheri,  Trigonia  pellati,  Exogyra  virgula. 

Aulacostephanus  Zones. 

Again  the  ammonite  fauna  of  these  beds  is  not  well  known,  but  on  the  basis  of  the 
recorded  species  and  the  associated  fauna,  several  correlations  may  be  suggested. 

Bed  1  corresponds  to  the  Elegans  Zone.  Gravesia  portlandica  (de  Loriol)  is 
probably  a  junior  synonym  of  G.  gigas  (Zieten).  Trigonia  pellati  and  Exogyra  virgula 
also  occur  in  this  zone  in  Dorset. 

Beds  2-4  probably  correspond  to  the  Scitulus,  Wheatleyensis  and  Hudlestoni  Zones. 
Pectinatites  (Virgatosphinctoides)  pringlei  (horizon  Ph.  1)  is  close  to  P.  (V.)  wheat- 
leyensis, and  is  probably,  therefore,  from  the  mid-Wheatleyensis  Zone. 

Bed  5  appears  to  correspond  to  the  upper  part  of  the  Pectinatus  Zone,  and  the 
Tour  Croi  Nodule  Bed  to  the  Rotundum  Zone. 

(c)  East  Greenland  (Milne  Land) 

The  Kimmeridgian  fauna  of  Greenland  were  the  subject  of  papers  by  Spath  (1935, 
1936).     He  described  therein  collections  made  on  expeditions  led  by  Dr.  Lauge  Koch. 
The  Upper  Kimmeridgian  succession  there  (Spath  1936  :  163)  is: 

Pavlovia  Beds.     150  ft. 
Pectinatites  Beds.     150  ft. 
Unfossiliferous  Shales.     120  ft. 
Band  of  crushed  Perisphinctids. 

The  Pectinatites  Beds  correspond  to  the  Pectinatus  Zone  of  Dorset.  Specific 
identity  is  established  with  Dorset  in  two  cases. 

The  band  of  crushed  Perisphinctids  yielded  three  specimens  which  Spath  identified 
tentatively  as  Subdichotomoceras? ,  Subplanites?  (Virgatosphinctoides?),  and  Sub- 
planites?  (Spath  1936,  pi.  1).  The  latter  two  specimens  appear  from  the  plate  to 
resemble  forms  from  the  Wheatleyensis  Zone,  and  are  probably  to  be  correlated  with 
this  zone.  The  unfossiliferous  beds  between  these  two  points  of  correlation  probably 
representing  the  Hudlestoni  Zone  of  Dorset.  No  fauna  to  be  correlated  with  Elegans 
or  Scitulus  Zones  is  recorded  from  East  Greenland. 

(d)  Southern  Germany  (Franconia) 

In  the  southern  part  of  Europe  the  ammonite  fauna  of  the  Upper  Jurassic  rocks 
becomes  markedly  different  from  that  of  North-west  Europe,  above  the  Lower 
Kimmeridgian.  To  these  rocks  equivalent  to  the  Upper  Kimmeridgian  and  Port- 
landian  Stages  of  North-west  Europe  the  stage  name  "  Tithonian  "  is  generally 
applied. 


74 


UPPER    KIMMERIDGE   CLAY   OF   DORSET 


LU 

*: 
en 
< 

>- 

CD 


LU 

ce 
en 

O 

u 


z 
lu 

LU 

I- 

LU 

CO 


cc 
cr 
O 
U 


u. 

o 


Z 
< 

rr 

_J 

iii 

1- 

I" 

2 

cc 
O 

Q. 

^ 

CC 

cc 

LU 

LU 
Q_ 

a. 

3 

o 

_l 

< 

cc 

LU 
CD 

li. 

o 


to 
o 

Uj 
CL 


LO 
Q 
LU 
CO 

o 
cc 

LU 
CO 

3 
LU 


Ui 

(^ 

<■ 

Cl 

"* 

CL 

-J 

CO 

u 

«1 

o 

LO 

z 

Ld 
Ll 

o 

I 

LO 

h- 

<r 

u 
z 
z 

LU 

CC 


m 
Q 

Ld 

CO 


CO 

I/) 

z 

LU 

1 

(/I 

_l 

3 

Q 

UI 

cc 

CO 

UI 

n 

Q. 

3 

LO 

Q 
UJ 
CO 

z 

LU 
U. 

O 

I 
z 

LU 

_l 

o 

LO 


LO      LO 
Z      LO 


O 


Q 


LU 

cc    < 

O     LU 
U     CL 

o 

Q    U 


h- 
LO 
LU 
O 
C9 
3 
(/) 


CC 


Q 

O 
2 


CD 

cc 

LU 
CO 

3 
UI 

z 


10 
Uj 

$ 

LO 
Ct 

5 


1-0 


o 
-J 

5 


Ld 

u_ 
O 

I 
u) 
i- 
cc 

LU 

z 
z 

UJ 

cc 


10 

Q 
LU 
CO 


CO 


LU 
L0 
00 

3 


(0   oo 

Uj    Ly 


•^ 

-^ 

n 

t—    t~. 

> 

3  ^t 

ri 

^  S 

K~ 

C:  C 

o 

o  o 

Lt 

Uj   Uj 

0.   Q. 

U)  -s 
Lu  LO 
i-~    Lu 


5  2 

§  a 

»—  to 

p  d 

Lu  l- 

a  ^ 

ct 


Lu 

9    LO 
!?LO 


LU 

X 

LO 

LO 

n 

z 

LU 

cc 

rfl 

LU 

O 

1 

oo 
Lu 
Q 

LO     O 

t  >  <o 

sap 

C    LO   Lu 
O    O   ,-J 

Lu    ►-   Lu 

o 
cc 


Fig.    12.     Correlations    between    the    Upper    Kimmeridgian    and    Lower    Portlandian   of 
Britain,  and  the  Lower  Tithonian  of  Franconia. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  75 

The  area  taken  herein,  as  representative  of  typical  Tithonian  rocks,  is  Franconia. 
I  have  recently  visited  this  area,  and  have  examined  large  ammonite  collections  made 
by  Dr.  A.  Zeiss  of  the  University  of  Erlangen,  with  whom  I  discussed  problems  of 
correlation  between  the  two  faunal  provinces.  The  results  of  these  discussions  were 
incorporated  in  a  joint  paper  (Cope  &  Zeiss  1964). 

The  most  firm  bases  for  correlation  are  to  be  found  between  the  basal  and  upper- 
most Upper  Kimmeridgian  and  basal  Portlandian,  and  their  Franconian  equivalents. 
In  the  Lower  Neuberg  Beds  specimens  of  Pavlovia  and  Zaraiskites  have  recently 
been  found.  This  gives  good  correlations  with  the  uppermost  Kimmeridgian  and 
basal  Portlandian  of  Dorset.  The  discoveries  are  particularly  important  since  it 
means  that  the  Kimmeridgian-Portlandian  boundary  can  be  traced  into  the  Tithonian 
faunal  province. 

At  the  base  of  the  Upper  Kimmeridgian  the  genus  Gravesia  occurs,  its  vertical 
range  in  Dorset  being  about  60  ft.  In  Franconia,  Gravesia  occurs  in  the  Moernsheim 
Beds  where  its  vertical  range  is  about  90  ft. 

Between  these  two  points  where  correlation  can  definitely  be  established,  the 
faunas  of  the  two  provinces  are  distinct.  In  Britain  species  of  Pectinatites  are  the 
commonest  ammonites,  and  in  Franconia  species  of  Subplanites,  Lithacoceras  and 
Pseudovirgatites.  As  stated  earlier  (p.  20)  Subplanites  and  Lithacoceras  do  not  occur 
in  Britain.  It  therefore  appears  that  the  apparent  similarity  between  the  two  faunas 
is  due  entirely  to  the  phenomenon  of  homeomorphy. 

It  is  remarkable  to  find  that  some  homeomorphs  seem  to  have  existed  contem- 
poraneously. Thus  some  ammonites  of  the  Pectinatites  [V irgatosphinctoid.es)  grandis 
group  are  very  close  to  undescribed  ammonites  from  the  Usseltal  Beds.  P.  (Virgato- 
sphinctoides)  reisiformis  has  a  very  similar  microconch  to  Subplanites  siliceus,  the 
apertural  modifications  of  the  two  forms  being  the  only  apparent  point  of  difference. 
P.  (Pectinatites)  inconsuetus  has  a  macroconch  almost  identical  in  appearance  to  an 
undescribed  species  of  Pseudovirgatites. 

Such  homeomorphs  cannot  provide  correlations,  but  it  has  been  found  that  their 
respective  stratigraphical  ranges  are  approximately  equal  in  some  cases.  It  appears 
that  direct  correlation  by  means  of  ammonites  is  not  possible  in  this  case,  and  the 
problem  is  unlikely  to  be  solved  until  an  area  is  discovered  where  an  overlap  of  the 
faunal  provinces  occurs. 

(e)  Russia  (Basin  of  the  Ural  and  Ilek  Rivers) 

The  Upper  Kimmeridgian  faunas  of  the  basin  of  the  Ural  and  Ilek  rivers  were 
described  by  Ilovaisky  &  Florensky  (1941).  The  specimens  they  described  came 
from  the  Vetlianka  Sandstone,  and  were  described  as  belonging  to  the  genus  Ilovaiskya 
Vialov  1940.  This  genus  was  regarded  by  Arkell  (1957)  as  a  junior  synonym  of 
Subplanites  Spath  1925. 

Although  several  of  the  forms  figured  by  Ilovaisky  &  Florensky  appear  very  similar 
to  British  species,  and  were  identified  as  such  by  Arkell  (1956  :  489-490),  identity 
even  at  generic  level  with  British  forms  cannot  be  established  on  the  basis  of  the 
published  plates.     None  of  the  ammonites  figured  by  Ilovaisky  &  Florensky  has  its 


76  UPPER    KIMMERIDGE    CLAY    OF    DORSET 

peristome  preserved,  and  thus  may  belong  equally  to  Subplanites  or  its  homeomorph 
Pectinatites.  As  no  specimens  similar  to  Lithacoceras  were  figured,  it  is  possible  that 
the  Russian  forms  belong  to  Pectinatites  rather  than  to  Subplanites.  The  collection 
of  material  with  peristomes  intact  is  essential,  however,  for  this  to  be  established 
with  certainty. 

This  problem  has  not  been  resolved  in  a  more  recent  paper  by  Michailov  (1964). 
He  figures  specimens  under  the  names  of  Subplanites  and  Pectinatites.  It  may  well 
be  that  in  parts  of  Russia  there  is  a  mixture  of  these  two  faunal  elements,  but  again 
the  absence  of  peristome-bearing  specimens  mean?  that  such  generic  placings  by 
Michailov  may  be  incorrect. 


VIII.      REFERENCES 

Arkell,  W.  J.      1933.     The  Jurassic  System  in  Great  Britain,      xii  +  681  pp.,  41  pis.,  97  figs. 

Oxford. 
1935-     On  the  Lower  Kimmeridgian  ammonite  genera  Pictonia,  Rasenia,  Aulacostephanus 

and  Ataxioceras.     Geol.  Mag.,  London,  72  :  246-257,  pi.   11. 
1935-1948.     A  monograph  of  the  ammonites  of  the  English  Corallian  Beds,  lxxxiv  +  420 

pp.,  78  pis.     Palaeontogr.  Soc.  (Monogr.),  London. 

1946.     Standard  of  the  European  Jurassic.     Bull.  geol.  Soc.  Am.,  New  York,  57  :  1-34. 

1947.     The  geology    of  the   country    around    Weymouth,    Swanage,    Corfe,    and   Lulworth. 

xii  +  386  pp.,  19  pis.  (Mem.  geol.  Surv.  U.K.,  London). 

1947a.     The  geology  of  Oxford,     vi  +  267  pp.,  6  pis.     Oxford. 

— —  1950.     A  classification  of  the  Jurassic  ammonites.     /.  Paleont.,  Tulsa,  24  :  354-364. 

1953-     Seven  new  genera  of  Jurassic  ammonites.     Geol.  Mag.,  London,  90  :  36-40,  pi.  1. 

1956.     Jurassic  Geology  of  the  World,     xv  +  806  pp.,  46  pis.     Edinburgh  &  London. 

Arkell,  W.  J.,  Kummel,  B.  &  Wright,  C.  W.     1957.     Mesozoic  Ammonoidea.     In  Treatise  on 

Invertebrate  Paleontology.  Part  L.  Mollusca  4.  Cephalopoda,  Ammonoidea.  (Ed.  by 
R.  C.  Moore).     L80-L437.     Kansas  &  New  York. 

Bailey,  E.  B.  &  Weir,  J.  1932.  Submarine  faulting  in  Kimmeridgian  times  in  East  Suther- 
land.    Trans.  R.  Soc.  Edin.,  57  :  429-467,  pis.  1-3. 

Basse,  E.  1930.  Contribution  a  l'etude  du  Jurassique  sup6rieur  (facies  corallien)  en  Ethiopie 
et  en  Arabie  meridionale.     Mem.  Soc.  geol.  Fr.,  Paris  (n.s.)  6  :  105-148,  pis.  4,  5. 

1952.     Cephalopodes.     In  Piveteau,  J.     1952.     Traite  de  Paleontologie,  2  :  461-755,  pis. 

1-24.     Paris. 

Berckhemer,  F.  &  Holder,  H.  1959.  Ammoniten  aus  dem  Oberen  Weissen  Jura  Siiddeutsch- 
lands.     Beih.  geol.  Jb.,  Hannover,  35  :  1-135,  27  pis. 

Blake,  J.  F.     1875.     On  the  Kimmeridge  Clay  of  England.     O.  Jl  geol.  Soc.  Lond.,  31  :  196-237. 

1880.     On  the  Portland  rocks  of  England.     Q.  Jl  geol.  Soc.  Lond.,  36  :  189-236. 

1 88 1.  On  the  correlation  of  the  Upper  Jurassic  rocks  of  England  with  those  of  the  Con- 
tinent.    Part  1.     The  Paris  Basin.     Q.  J I  geol.  Soc.  Lond.,  37  :  497-587. 

Blanchet,  F.  1923.  La  faune  du  Tithonique  inferieur  des  regions  subalpines  et  ses  rapports 
avec  celle  du  Jura  franconien.     Bull.  Soc.  geol.  Fr.,  Paris  (4)  23  :  70-80. 

Buckman,  S.  S.     1922-1923.     Type  Ammonites,  4  :  1-67,  pis.  268A-422.     London. 

1923-1925.     Type  Ammonites,  5  :  1-88,  pis.  423-576.     London. 

1925-1927.     Type  Ammonites,  6  :  1-61,  pis.  528A-717B.     London. 

■ 1927-1930.     Type   Ammonites ,    7  :  1-78,    pis.    31A-773A.     With    editorial   notes   by   A. 

Morley  Davies.     London  &  Thame. 

Buckman,  S.  S.  &  Bather,  F.  A.  1894.  Can  the  sexes  in  ammonites  be  distinguished?  Nat. 
Sci.,  London,  4  :  427-432. 


UPPER    KIMMERIDGE    CLAY    OF    DORSET  77 

Callomon,   J.  H.     1955.     The  ammonite  succession  the  Lower  Oxford  Clay  and  Kellaways 

Beds  at  Kirtlington,  Oxfordshire,  and  the  zones  of  the  Callovian  Stage.     Phil.  Trans.  R. 

Soc,  London,  239,  B  :  215-264,  pis.  2,  3. 
1957-     Field  meeting  in  the  Oxford  Clay  of  Calvert  and  Woodham  Brick  Pits,  Buckingham- 
shire.    Proc.  Geol.  Ass.,  London,  68  :  61-64. 
1961.     The    Jurassic    System    in    East    Greenland.     In    Raasch,    G.    O.     1961    (Editor). 

Geology  of  the  Arctic  :  258-268.     Toronto. 
1963.     Sexual  dimorphism  in  Jurassic  ammonites.     Trans.  Leicester  lit.  phil.  Soc,  57  :  21- 

56,  pi.  1. 
Chatwin,  C.  P.  &  Pringle,  J.      1922.     The  zones  of  the  Kimmeridge  and  Portland  rocks  at 

Swindon.     Summ.  Progr.  Geol.  Surv.  Lond.,  1921  :  162-168. 
Cope,  J.  C.  W.,  Sarjeant,  W.  A.  S.,  Spalding,  D.  A.  E.  &  Zeiss,  A.     1964.     The  Kimmeridgian- 

Portlandian  boundary.     In  Colloque  du  Jurassique,  Luxembourg  1962.      Volume  des  comptes 

rendus  et  memoires,  Luxembourg  :  933-936. 
Cope,  J.  C.  W.  &  Zeiss,  A.     1964.     Zur  Parallelisierung  den  englischen  Oberkimmerdige  mit 

dem  frankischen  Untertithon  (Malm  Zeta).     Geol.  Bl.  Nordost-Bayern,  Erlangen,  14  :  5-14. 
Dacque,  E.     1905.     Beitrage  zur  Geologie  des  Somalilandes.     2  Oberer  Jura.     Beitr.  Palaont. 

Geol.  Ost.-Ung.,  Wien,  17  :  119-160,  pis.  14-18. 
Damon,  R.     1884.     Geology  of  Weymouth,  Portland,  and  the  coast  of  Dorsetshire,  from  Swanage  to 

Bridport-on-the-Sea:   with  natural  history  and  archaeological  notes.     2nd  (revised)  ed.  xii  -f 

250  pp.     London. 

1888.     A  supplement  to  the  Geology  of  Weymouth.     2nd.  (revised)  ed.     20  pis.     London. 

Donovan,  D.  T.     1957.     The  Jurassic  and  Cretaceous  Systems  in  East  Greenland.     Meddr. 

Gmnland,  Copenhagen,  155  :  1-2 14,  pis.  1-4. 
Donze,  P.  &  Enay,  R.     1 961.     Les  cephalopodes  du  Tithonique  inferieur  de  la  Croix-de-Saint- 

Concors  pres  Chamb6ry  (Savoie).     Trav.  Lab.  Geol.  Univ.  Lyon  (n.s.)  7  :  1-236,  pis.  1-22. 
Douville,  R.     1910.     Un  Virgatites  du  Caucase  occidental ;   origine  m6diterran6enne  du  genre  ; 

Ataxioceras,  Pseudovirgatites  et  Virgatosphinctes.     Bull.  Soc.  geol.  Fr.,  Paris  (4)  10  :  730-739, 

1  fig. 
Dumortier,  E.  &  Fontannes,  F.     1876.     Description  des  ammonites  de  la  zone  a  Ammonites 

tenuilobatus  de  Crussol  (Ardeche)  et  de  quelques  autres  fossiles  jurassiques  nouveaux  ou  peu 

connus.     159  pp.,  19  pis.     Paris. 
Dutertre,  A.  P.     1926.     Contribution  a  l'6tude  paleontologique  du  Portlandien  du  Boulonnais. 

Annls  Soc.  geol.  N.,  Lille,  51  :  240-249. 
1927.     Remarques  sur  la  position  stratigraphique  de  quelques  ammonites  du  Portlandien 

moyen  du  Boulonnais.     C.  r.  somm.  Seanc.  Soc.  geol.  Fr.,  Paris  (4)  27  :  36-37. 
Favre,  E.     1877.     La  zone  a  Ammonites  acanthicus  dans  les  Alpes  de  la  Suisse  et  de  la  Savoie. 

Abh.  schweiz.  palaont.  Ges.,  Zurich,  4  :  1-112,  pis.  1-9. 
Frebold,  H.     1928.     Das  Festungsprofil  auf  Spitzbergen,   Jura  und  Kreide.     Skr.  Svalbard 

Ishavet,  Oslo,  19  :  183-201. 
Geyer,  O.  F.     1961.     tJber  die  alteste  virgatipartite  Berippung  der  Perisphinctidae  (Cephalo- 
poda).    Palaont.  Z.,  Stuttgart,  35  :  90-94. 
1961a.     Monographie    der    Perisphinctidae    des    unteren    Unter-Kimeridgium    (Weisser 

Juray,   Badenerschichten)    im    siiddeutschen    Jura.     Palaeontographica,     Stuttgart,     117, 

A  :  1-157,  pis.  1-22. 
Glangeaud,  P.     1897.     Sur  la  forme  de  l'ouverture  de  quelques  ammonites.     Bull.  Soc.  geol. 

Fr.,  Paris  (3)  25  :  99-107. 
Hahn,  W.     1963.     Die  Gattung  Gravesia  Salfeld  im  Oberjura  Mittel — und  Nordwesteuropas. 

Palaeontographica,  Stuttgart,  122,  A  :  90-110,  pis.  9-13. 
Holder,  H.     1961.     Uber  den  Ammonites  ulmensis  Oppel.  Jber.  Mitt,  oberrhein.  geol.   Ver., 

Stuttgart,  43  :  11 3-1 17,  pi.  5. 
Hudleston,  W.  H.      1896.     In  Hudleston,  W.  H.  Mansel,  O.  L.  &  Monckton,  H.  W.  Excur- 
sion to  Swanage,  Corfe  Castle,  Kimeridge,  etc.  Proc.  Geol.  Ass.,  London,  14,  312-323. 


78  UPPER    KIMMERIDGE   CLAY    OF   DORSET 

Ilovaisky,  D.  &  Florensky,  K.  P.     1941.     Les  ammonites  du  Jurassique  superieur  des  bassins 

des  rivieres  Oural  et  Ilek.  Contrib.  Connaiss.  ge'ol.  U.R.S.S.,  Soc.  Nat.  Moscou  (n.s.)  5  :  1-195, 

pis.  1-28.     (In  Russian). 
International  Commission  for  Zoological  Nomenclature.     1961.     International  Code  of 

Zoological  Nomenclature  adopted  by  the  15th.  International  Congress  of  Zoology,     xvii  -+-  176 

pp.     London. 
Kerr,  J.  C.      1895.     On  some  points  of  the  anatomy  of  Nautilus  pompilius.     Proc.  zool.  Soc. 

Land.,  65  :  1-3. 
Lamplugh,  G.  W.,  Kitchin,  F.  L.  &  Pringle,  J.     1923.     The  concealed  Mesozoic  rocks  in  Kent. 

iv  +  248  pp.,  2  pis.  (Mem.  geol.  Surv.  U.K.,  London). 
Lane,  F.  W.     1957.     The  Kingdom  of  the  Octopus,     xx  +  287  pp.,  48  pis.,  London. 
Lewinsky,   J.     1922-1923.     Monographie  geologique  et  paleontologique  du  Bononien  de  la 

Pologne.     Mem.  Soc.  geol.  Fr.,  Paris,  24-25  :  1-108,  pis.  1-11. 
Linsley,  E.  G.,  Mayr,   E.   &  Usinger,   R.  L.     1953-     Methods  and  principles  of  systematic 

zoology,     ix  +  328  pp.,  14  tables,  45  figs.,  New  York. 
Makowski,    H.     1962.     The  problem   of  sexual  dimorphism   in   ammonites.     Palaeont.  pol., 

Warsaw,  12  :  i-viii  +  1-92,  pis.  1-20. 
Medcof,  J.  C.     1955.     Day  and  night  characteristics  of  spatfall  and  of  behaviour  of  oyster 

larvae.     J .  Fish.  Res.  Bd.  Can.,  Ottawa,   12  :  270-286. 
Mikhailov,    N.    P.     1964.     Boreal   late    Jurassic   cephalopods.     Trudy,   geol.    Inst.   Leningr., 

Moscow,  107  :  1-88,  pis.  1-20.     (In  Russian). 
Munier-Chalmas,  E.  C.  P.  A.     1892.     Sur  la  possibilit6  d'admettre  un  dimorphism  sexuel  chez 

les  ammonitides.     C.  r.  somm.  Se'anc.  Soc.  geol.  Fr.,  Paris  (3)  20  :  170-174. 
Neaverson,  E.     1924.     The  zonal  nomenclature  of  the  Upper  Kimmeridge  Clay.     Geol.  Mag., 

London,  61  :  146-15 1. 
1925.     Ammonites  from  the  Upper  Kimmeridge  Clay.     Pap.  geol.  Dep.    Univ.  L'pool., 

1  :  1-52,  pis.  1-4. 
Orbigny,    A.    D'.     1842-1851.     Paleontologie    Francaise,    Terrains    Jurassiques,    1.   642    pp., 

234  pis.     Paris. 
Oppel,    A.     1862-1863.     Uber   jurassische    Cephalopoden.     Pal.    Mitt.    Mus.    k.    Bayer.    St., 

Stuttgart,  1  :  127-266,  pis.  40-74. 
Pavlow,  A.  P.     1896.     On  the  classification  of  the  strata  between  the  Kimmeridgian  and  Aptian. 

Q.  J I  geol.  Soc.  Lond.,  52  :  542-555,  pi.  27. 
Pompeckj,    J.     1894.     Uber   Ammonoideen   mit   anormaler   Wohnkammer.     Jh.    Ver.   vaterl. 

Naturk.  Wurtt.,  Stuttgart,  49  :  220-290,  pi.  4. 
Pruvost,    P.     1924.     Les    subdivisions    du    Portlandien    boulonnais    d'apres    les    ammonites. 

Annls  Soc.  geol.  N.,  Lille,  49  :  187-215,  pi.  2. 
Pruvost,  P.  &  Pringle,  J.      1924.     A  synopsis  of  the  geology  of  the  Boulonnais,  including  a 

correlation  of  the  Mesozoic  rocks  with  those  of  England.     Proc.  Geol.  Ass.,  London,  35  : 

26-67. 

1924a.     Observations  sur  la  serie  portlandienne  du  Boulonnais.     C.  r.  hebd.  Se'anc.  Acad. 

Sci.  Paris,  177  :  398-401. 

Ouenstedt,  F.  A.     1887-1888.     Die  Ammoniten  des  Schwdbischen  Jura.     3,  Der  Weisse  Jura  : 

817-1140     Atlas  pis.  91-126.     Stuttgart. 
Roll,  A.     1932.     Stratigraphischer  Vergleich  zwischen  nordwesteuropaischen  und  siiddeutschen 

oberen  Malm.     Neues  Jb.  Miner.  Geol.  Paldont.  BeilBd.,  Stuttgart,  68,  B  :  179-198. 

1933-     Uber    den    oberen    Malm    des    siidwestlichen    Frankenalb.     Zentbl.    Miner.    Geol. 

Paldont.,  Stuttgart,  10,  B  :  553-564. 

Rollier,  L.     1913.     Sur  quelques  ammonoides  jurassiques  et  leur  dimorphisme  sexuel.     Archs. 

Sci.  phys.  nat.,  Geneve  (4)  35  :  263-268. 
Salfeld,  H.     1913.     Certain  Upper  Jurassic  strata  of  England.    Q.  J I  geol.  Soc.  Lond.,  69  :  423- 

432,  pis.  41,  42. 

1914-     Die   Gliederung   des   oberen   Jura  in   Nordwesteuropa.     Neues   Jb.   Miner.   Geol. 

Paldont.,  Stuttgart,  37  :  125-246,  2  figs. 


UPPER    KIMMERIDGE   CLAY    OF   DORSET  79 

Sauvage,   H.   E.     191 1.     Sur  quelques  ammonites  du   Jurassique  superieur  du   Boulonnais. 

Bull.  Soc.  ge'ol.  Fr.,  Paris  (4)  11  :  455-462. 
Schenk,    E.    T.    &   McMasters,    J.    H.      1948.     Procedure   in    Taxonomy.     2nd    (revised)    ed. 

vii  +  93  pp.     Stanford  &  London. 
Schindewolf,  O.  H.      1925.     Entwurf  einer  Systematik  der  Perisphincten.     Neues  Jb.  Miner. 

Geol.  Palaont.  BeilBd.,  Stuttgart,  52,  B  :  309-343. 

1926.     Zur   Systematik   der   Perisphincten.     Neues   Jb.   Miner.    Geol.   Palaont.    BeilBd., 

Stuttgart,  55,  A  :  495-517,  pi.  19. 

Schneid,  T.     1914-1915.     Die  Geologie  der  frankischen  Alb  zwischen  Eichstatt  und  Neuberg 
a.D.  Geogn.  Jh.,  Miinchen,  27-28  :  59-229,  pis.  1-9. 

1915.     Die  Ammonitenfauna  der  Obertithonisches  Kalke  von  Neuberg  a.D.  Geol.  palaont. 

Abh.,  Jena  (N.F.)  13  :  305-416,  pis.  17-29. 

Seilacher,  A.     i960.     Epizoans  as  a  key  to  ammonoid  ecology.     /.  Paleont.,  Tulsa,  34  :  189- 

193- 
Spath,    L.    F.     1923.     On   ammonites   from    New   Zealand.     Appendix    to   Trechman  :     The 
Jurassic  of  New  Zealand.     Q.  Jl  geol.  Soc.  Lond.,  79  :  286-312,  pis.  12-18. 

1925.     The  collection  of  fossils  and  rocks  from  Somaliland  made  by  Messrs.  Wyllie  and 

Smellie,  7.     Ammonites  and  aptychi.     Monogr.  geol.  Dep.  Hunter.  Mus.,  Glasgow,  1  :  in- 
164,  pis.  14,  15. 

1927-1933.      Revision    of    the    Jurassic    cephalopod    fauna    of    Kachh    (Cutch),     1-6. 

Palaeont.  indica,  Calcutta  (n.s.)  9,  2.     vii  +  955  pp.,  130  pis. 

J935-     The  Upper  Jurassic  invertebrate  faunas  of  Cape  Leslie,  Milne  Land,  1.     Oxfordian 

and  Lower  Kimmeridgian.     Meddr.  Gronland,  Copenhagen,  99,  2  :  1-82,  pis.  1-15. 

1936.     The  Upper  Jurassic  invertebrate  faunas  of  Cape  Leslie,  Milne  Land,  2.     Upper 

Kimmeridgian  and  Portlandian.     Meddr.  Gronland,  Copenhagen,  99,  3  :  1-180,  pis.  1-50. 

Stenzel,  H.  B.     1964.     Living  Nautilus.     In  Treatise  on  Invertebrate  Palaeontology .     Part  K. 

Mollusca  3.     (Ed.  by  R.  C.  Moore).     K59-K93.     Kansas  &  New  York. 
Strahan,  A.     1898.     Geology  of  the  Isle  of  Purbeck.     xi  +  278  pp.,  11  pis.,  183  figs.     (Mem.  geol. 

Surv.  U.K.,  London). 
Toucas,  M.  A.     1889.     Nouvelles  observations  sur  le  Jurassique  superieur  de  l'Ardeche.     Bull. 

Soc.  geol.  Fr.,  Paris  (3)  17  :  729-742. 
1890.     Etude  de  la  faune  des  couches  tithoniques  de  l'Ardeche.     Bull.  Soc.  geol.  Fr.,  Paris 

(3)  18  :  560-629. 
Vetters,  H.     1905.     Die  Fauna  der  Juraklippen  zwischen  Donau  und  Thaya.     Beitr.  Palaont. 

Geol.  Ost.-Ung.,  Wien,  17  :  223-259,  pis.  21,  22. 
Westermann,  G.  E.  G.     1964.     Sexual-Dimorphismus  bei  Ammonoideen  und  seine  Bedeutung 

fur  die  Taxionomie  der  Otoitidae.     Palaeontographica ,  Stuttgart,  124,  A  :  33-73,  pis.  6-9. 
Willey,    A.      1897.     The    oviposition    of    Nautilus    macromphalus .     Proc.    R.    Soc,    London, 

60:   467-471. 
1897a.     Adhesive  tentacles  of  Nautilus,  with  some  notes  on  its  pericardium  and  spermato- 

phores.     Q.  Jl  microsc.  Sci.,  London,  40  :  207-209,  pi.  2. 
—  1899.     Some  zoological  results  of  a  voyage  to  Melanesia  during  the  years   1894-1897. 

Proc.  Camb.  phil.  Soc,  London,  9,  8  :  397-400. 

1902.     Zoological  results  based  on  material  from  New  Britain,  New  Guinea,  Loyalty  Islands 

and  elsewhere.     6.  Contribution  to  the  natural  history  of  the  Pearly  Nautilus  :  691-826,  pis. 
75-83.     Cambridge. 

Yonge,  C.  M.     i960.     Oysters,     xiv  +  209  pp.,  17  pis.     London. 

Ziegler,     B.      1962.     Die    Ammoniten-Gattung    Aulacostephanus    im    Oberjura.     Palaeonto- 
graphica, Stuttgart,  119,  A  :  1-172,  pis.  1-22. 


PLATES 

The  photographs  are  by  Mr.  S.  P.  Osborn  of  the  Geology  Department,  University 
College,  Swansea. 

All  the  specimens  were  whitened  with  ammonium  chloride  prior  to  photographing. 


PLATE  i 
Fig.  i.     Gravesia  gigas  (Zieten).  C.  73390,   xo-45,  45  feet  below  Yellow  Ledge  Stone  Band. 
Fig.  2.     Gravesia  cf.  gravesiana  (d'Orbigny).  C.  73391,  x  1,  8  feet  below  Yellow  Ledge  Stone 
Band. 


Bull.  Br.  Mus.  not.  Hist.  (Geol.)  15,  1 


PLATE    1 


GEOL.   15,    I. 


PLATE  2 

Fig.  i.  Pectinatites  (Arkellites)  primitivus  sp.  nov.  Holotype  (macroconch),  C. 73392, 
X  i,  25  feet  below  Yellow  Ledge  Stone  Band. 

Fig.  2.  Pectinatites  (Arkellites)  primitivus  sp.  nov.  Paratype  (microconch),  C. 73395, 
X  i,  55  feet  below  Yellow  Ledge  Stone  Band. 

Fig.  3.  Pectinatites  (Arkellites)  hudlestoni  sp.  nov.  Microconch,  C. 73402,  x  1-5,  ventral 
view  showing  possible  points  of  shedding  of  horns.     13  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    2 


PLATE  3 

Fig.  i.  Pectinatites  (Arkellites)  primitivus  sp.  nov.  Paratype  (macroconch) ,  C. 73393, 
X  1,  25  feet  below  Yellow  Ledge  Stone  Band. 

Fig.  2.  Pectinatites  (Arkellites)  primitivus  sp.  nov.  Paratype  (microconch) ,  C. 73394. 
X  1,  25  feet  below  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  I 


PLATE    3 


PLATE  5 

Fig.  i.  Pectinatites  (Arkellites)  cuddlensis  sp.  nov.  Paratype  (microconch),  C. 73397, 
X  1,  25  feet  above  Yellow  Ledge  Stone  Band. 

Fig.  2.  Pectinatites  (Arkellites)  damoni  sp.  nov.  Paratype  (microconch),  C. 73400,  x  1, 
25  feet  above  Yellow  Ledge  Stone  Band. 

Fig.  3.  Pectinatites  (Arkellites)  damoni  sp.  nov.  Paratype  (microconch),  C. 73401,  x  1, 
25  feet  above  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  not.  Hist.  (Geol.)  15,  1 


PLATE    5 


PLATE  6 

Fig.  i.  Pectinatites  (Arkellites)  damoni  sp.  now  Paratype  (macroconch),  C. 73399,  XI, 
27  feet  above  Yellow  Ledge  Stone  Band. 

Fig.  2.  Pectinatites  {Arkellites)  damoni  sp.  nov.  Holotype  (macroconch),  C. 73398,  x  1, 
25  feet  above  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE   6 


PLATE  7 
Pectinatites  (Arkellites)  hudlestoni  sp.  nov.    Holotype  (macroconch) ,  C. 73403,    xo-85, 
13  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    7 


PLATE  8 

Fig.  ia.  Pectinatites  (Virgatosphinctoides)  elegans  sp.  nov.  Paratype  (microconch) , 
0.73406,    x  1,  20  feet  below  Yellow  Ledge  Stone  Band. 

Fig.  ib.     Reverse  side  of  specimen  in  Fig.  ia,  showing  detail  of  the  peristomal  inflation,      x  I. 

Fig.  2.  Pectinatites  (Arkellites)  hudlestoni  sp.  nov.  Paratype  (microconch),  0.73404, 
X  1,  13  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  not.  Hist.  (Geol.)  15,  1 


PLATE    8 


PLATE  9 
Pectinatites  (Virgatosphinctoides)  elegaris  sp.  nov.    Holotype  (macroconch),  C. 73405, 
yo-95,  18  feet  below  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE   9 


GEOL.  15,   I. 


PLATE  10 
Fig.    i.     Pectinatites   (Virgatosphinctoides)    elegans   corniger 

(macroconch),  C. 73407,   X  1,  5  feet  below  Yellow  Ledge  Stone  Band. 

Fig.    2.     Pectinatites   (Virgatosphinctoides)   elegans   corniger 
(microconch),  C. 73409,    x  1,  8  feet  below  Yellow  Ledge  Stone  Band. 

Fig.    3.     Pectinatites   (Virgatosphinctoides)    elegans   corniger 

(microconch),  C. 73408,   xi,  5  feet  below  Yellow  Ledge  Stone  Band. 


subsp. 

nov.     Holotype 

subsp. 

nov.    Paratype 

subsp. 

nov.    Paratype 

Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    10 


.g^fHv* 


PLATE   i  i 

Fig.  i.  Pectinatites  (Virgatosphinctoides)  scitulus  sp.  nov.  Holotype  (macroconch), 
C. 7341 1,   xo-85,  24  feet  above  Yellow  Ledge  Stone  Band. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  scitulus  sp.  nov.  Paratype  (microconch) , 
C. 73412,   xo-85,  25  feet  above  Yellow  Ledge  Stone  Band. 

Fig.  3.  Pectinatites  (Virgatosphinctoides)  scitulus  sp.  nov.  Paratype  (microconch), 
C. 73413,   xo-85,  15  feet  above  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    11 


PLATE  12 

Fig.  i.  Pectinatites  (Virgatosphinctoides)  decorosus  sp.  nov.  Holotype  (macroconch), 
C. 73414,   X  1,  15  feet  above  Yellow  Ledge  Stone  Band. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  decorosus  sp.  nov.  Paratype  (microconch), 
C. 73415,   X  1,  15  feet  above  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    12 


PLATE  13 
Pectinatites   (Virgatosphinctoides)   major  sp.   nov.    Holotype   (macroconch),   C. 73410, 
X  0-55,  6  feet  below  Yellow  Ledge  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    13 


Fig.  i 
•73432, 
Fig.    2 

•73433- 
Fig.   3 

•73434. 


PLATE  14 
Pectinatites  ( Virgatosphinctoides)  clavelli 

X  0-7,  8  feet  above  Grey  Ledge  Stone  Band. 

Pectinatites   (Virgatosphinctoides)   clavelli 
x  0-7,  3  feet  above  Grey  Ledge  Stone  Band. 

Pectinatites   ( Virgatosphinctoides)   clavelli 

X07,  3  feet  above  Grey  Ledge  Stone  Band. 


sp.  nov.  Holotype  (macroconch), 
sp.  nov.  Paratype  (microconch) , 
sp.   nov.    Paratype    (microconch), 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    14 


PLATE   15 

Fig.  1.  Pectinatites  (Virgatosphinctoides)  smedmorensis  sp.  nov.  Holotype  (macro- 
conch),  C.  73430,   x  1,  22  feet  below  Blackstone. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  smedmorensis  sp.  nov.  Paratype  (micro- 
conch),  C. 73431,  x  1,  22  feet  below  Blackstone.  The  postulated  original  shell  outline  indicated 
by  broken  lines. 

Fig.  3.  Pectinatites  (Virgatosphinctoides)  grandis  (Neaverson).  Microconch,  C. 73421, 
xo-6,  17  feet  below  Blackstone. 


Bull.  By.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    t5 


PLATE  16 
Pectinatites  (Virgatosphinctoides)  laticostatus  sp.  nov.  Holotype  (macroconch) ,  C. 73416, 
X  0-65,  19  feet  below  Blackstone. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    16 


PLATE   iS 
Pectinatites  (Virgatosphinctoides)  grandis   (Neaverson).  Macroconch,  C.734Z°.    >•  °"45. 
4  feet  below  Blackstone. 


Bull.  Br.  Mus.  nal.  Hist.  (Geol.)  ig,  i 


PLATE    ii 


GEOL.   15,   I. 


8§ 


PLATE  19 
Pectinatites  (Virgatosphinctoides)  grandis  acceleratus  subsp.  nov.    Holotype  (macro- 
conch),  C. 73422,   Xo-45,  13  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    19 


*V  Is 


- 


PLATE  20 

Fig.  i.  Pectinatites  (Virgatosphinctoides)  woodwardi  (Neaverson).  Macroconch, 
C. 73423,   x  1,  15  feet  below  Blackstone. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  woodwardi  (Neaverson).  Microconch, 
C. 73424,  X  1,  15  feet  below  Blackstone.  The  postulated  original  extent  of  the  horn  indicated 
by  broken  line. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    20 


PLATE  21 

Fig.  i.  Pectinatites  (Virgatosphinctoides)  wheatleyensis  (Neaverson) .  Macroconch, 
C. 73425,   x  1,  12  feet  below  Blackstone. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  wheatleyensis  (Neaverson).  Microconch, 
C. 73426,   x  1,  12  feet  below  Blackstone. 

Fig.  3.  Pectinatites  (Virgatosphinctoides)  wheatleyensis  (Neaverson).  Microconch, 
C.  73427,   x  1,  15  feet  below  Blackstone. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    21 


PLATE  22 


Pectinatites  (Virgatosphinctoides)  reisiformis  sp.  nov.  Holotype  (macroconch),  C. 73435, 
xi,  13  feet  above  Rope  Lake  Head  Stone  Band. 


Btrfl.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    22 


PLATE  23 

Fig.  1.  Pectinatites  (Virgatosphinctoid.es)  reisiformis  densicostatus  subsp.  nov. 
Holotype  (macroconch) ,  C. 73437,    xo-85,  13  feet  above  Rope  Lake  Head  Stone  Band. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  reisiformis  densicostatus  subsp.  nov. 
Paratype  (microconch) ,  C.  73438,   xo-85,  13  feet  above  Rope  Lake  Head  Stone  Band. 

Fig.  3.  Pectinatites  (Virgatosphinctoides)  reisiformis  sp.  nov.  Paratype  (microconch), 
C. 73436,   xo-85,  13  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    23 


PLATE  24 

Fig.  1.  Pectinatites  (Virgatosphinctoides)  wheatleyensis  minor  subsp.  nov.  Holotype 
(macroconch),  C. 73429,   X085,  17  feet  below  Blackstone. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  reisiformis  densicostatus  subsp.  nov. 
Intersex,  C.  73439,  x  0-85,  13  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  r 


PLATE    24 


PLATE  25 

Fig.  1.  Pectinatites  (Virgatosphinctoides)  donovani  sp.  nov.  Holotype  (macroconch), 
C.  73441,   xo-85,  30  feet  below  Basalt  Stone  Band. 

Fig.  2.  Pectinatites  (Virgatosphinctoides)  donovani  sp.  nov.  Paratype  (microconch) , 
C. 73442,   xo-85,  36  feet  below  Basalt  Stone  Band. 

Fig.  3.  Pectinatites  (Pectinatites)  cornutifer  (Buckman).  Microconch,  C. 73451,  x  1, 
6  feet  above  Freshwater  Steps  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  i 


PLATE   25 


PLATE  26 

Fig.  1.  Pectinatites  (Pectinatites)  eastlecottensis  (Salfeld).  Microconch,  C. 73449,  x  1, 
10  feet  above  Middle  White  Stone  Band. 

Fig.  2.  Pectinatites  (Pectinatites)  cornutifer  (Buckman).  Microconch,  C. 73450,  x  1, 
6  feet  below  Freshwater  Steps  Stone  Band. 

Fig.  3.  Pectinatites  (Virgatosphinctoides)  abbreviatus  sp.  nov.  Holotype  (macroconch) , 
C. 73440,   x  1,  20  feet  above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    26 


M         ■ 


PLATE  27 

Fig.  1.  Pectinatites  (Virgatosphinctoides)  encombensis  sp.  nov.  Holotype  (microconch), 
C.  73444,   X  1,  21  feet  below  White  Stone  Band. 

Fig.  2.  Pectinatites  {Virgatosphinctoides)  wheatleyensis  delicatulus  (Neaverson). 
Macroconch,  C. 73428,   x  1,  4  feet  below  Blackstone. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE   27 


2**S 


GEOL.    1.5,1. 


PLATE  28 
Pectinatites    (Virgatosphinctoides)    encombensis    sp.    nov.     Paratype    (macroconch), 
C. 73445,   X075,  33  feet  below  White  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    28 


PLATE  29 
Pectinatites  (Virgatosphinctoides)   magnimasculus  sp.   nov.    Holotype   (microconch), 
C. 73443,  X  i,  21  feet  below  White  Stone  Band.     The  postulated  rib  density  of  missing  parts  of 
the  shell  indicated  by  broken  lines. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE    29 


PLATE  30 

Fig.  1.  Pectinatites  (Pectinatites)  inconsuetus  sp.  nov.  Paratype  (micrconch),  C. 73447, 
X  1,  10  feet  above  Middle  White  Stone  Band. 

Fig.  1.  Pectinatites  (Pectinatites)  inconsuetus  sp.  nov.  Holotype  (macroconch) ,  C.  73446, 
xo-g,  10  feet  above  Middle  White  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


1'LATE    30 


"*£  I  ft  *     '  '     "V 


PLATE  31 
Pectinatites  (Pectinatites)  groenlandicus  (Spath).  Macroconch,  C. 73448,    xo^,  10  feet 
above  Middle  White  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE  31 


Vfc     "tew. 


PLATE  32 

Fig.  1.  Pectinatites  (Pectinatites)  naso  (Buckman).  Macroconch,  0.73452,  x  1,  10  feet 
above  Freshwater  Steps  Stone  Band. 

Fig.  2.  Pectinatites  {Pectinatites)  naso  (Buckman).  Microconch,  C. 73453,  x  1,  10  feet 
above  Rope  Lake  Head  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE   32 


PLATE  33 
Pavlovia  (Paravirgatites)  cf.  paravirgatus  (Buckman).  C.  73454,  x  11,  10 feet  above  Fresh- 
water Steps  Stone  Band. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  1 


PLATE   33 


PRIKTBD  IN  GREAT  BRITAIN 
BY  ADLARD  &  SON  LIMITED 
BARTHOLOMEW   PRESS,   DORKING 


THE  CORRELATION  AND 

TRILOBITE  FAUNA  OF  THE  BEDINAN 

FORMATION  (ORDOVICIAN)  IN 

SOUTH-EASTERN  TURKEY 


%. 


W.  T.  DEAN 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  2 

LONDON:   1967 


12  JU 

THE    CORRELATION    AND    TRILOBITE    FAUNA 

OF  THE  BEDINAN  FORMATION  (ORDOVICIAN) 

IN  SOUTH-EASTERN  TURKEY 


BY 

WILLIAM  THORNTON  DEAN,  D.Sc. 


1- 


Pp.  81-123  ;   10  Plates  ;  4  Text-figures 


BULLETIN  OF 

THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 

GEOLOGY  Vol.  15  No.  2 

LONDON:  1967 


THE     BULLETIN    OF    THE    BRITISH    MUSEUM 

(NATURAL  history),  instituted  in  1949,  is 
issued  in  five  series  corresponding  to  the  Departments 
of  the  Museum,  and  an  Historical  series. 

Parts  will  appear  at  irregidar  intervals  as  they  become 
ready.  Volumes  will  contain  about  three  or  four 
hundred  pages,  and  will  not  necessarily  be  completed 
within  one  calendar  year. 

In  1965  a  separate  supplementary  series  of  longer 
papers  was  instituted,  numbered  serially  for  each 
Department. 

This  paper  is  Vol.  15,  No.  2  of  the  Geological 
(Palaeontological)  series.  The  abbreviated  titles  of 
periodicals  cited  follow  those  of  the  World  List  of 
Scientific  Periodicals. 


World  List  abbreviation: 
Bull.  Br.  Mus.  nat.  Hist.  (Geo!.). 


1  Trustees  of  the  British  Museum  (Natural  History)  1967 


TRUSTEES    OF 
THE    BRITISH    MUSEUM    (NATURAL    HISTORY) 

Issued  13  June,  1967  Price  £2 


THE    CORRELATION    AND    TRILOBITE    FAUNA 

OF  THE  BEDINAN  FORMATION  (ORDOVICIAN) 

IN  SOUTH-EASTERN  TURKEY 

By  W.  T.  DEAN 

MS  accepted  November,  30,  1966 


CONTENTS 

I.  Introduction  and  Acknowledgments 

II.  The  Succession  at  Bedinan        .... 

III.  The  Succession  near  Sosink       .... 

IV.  Age  and  Relationships  of  the  Bedinan  Faunas 
V.  Systematic  descriptions    ..... 

Ampyx  nitidus  sp.  nov.  .... 

Marrolithoides  orthogonius  sp.  nov. 
Marrolithoides  laticirrus  sp.  nov. 
Cryptolithus?  inferus  sp.  nov. 
Cryptolithus!  bedinanensis  sp.  nov. 
Dionide  formosa  (Barrande)  anatolica  subsp.  nov. 
Dalmanitina  proaeva  proaeva  (Emmrich)  . 
Kloucekia  phillipsii  (Barrande)  euroa  subsp.  nov. 
Cheirurid  gen.  et  sp.  indet.  .... 
Neseuretus  (Neseuretinus)  turcicus  subgen.  et  sp.  nov 
Brongniartella  levis  sp.  nov.  .... 
Platycoryphe?  sp.  . 
Colpocoryphe  sp.  . 
Selenopeltis  inermis  (Beyrich)  angusticeps  subsp.  nov 
Asaphid  gen.  et  sp.  indet.  .... 
VI.     References       ....... 

SYNOPSIS 


85 
89 
91 
93 
93 
96 

99 
102 
104 
109 
112 

"3 

114 

"5 
118 
120 
120 
121 
121 
122 


The  strata  of  the  Bedinan  Formation  are  described  from  the  type  region  between  Derik  and 
Mardin,  south-eastern  Turkey.  The  rocks,  which  rest  unconformably  upon  the  Sosink  Forma- 
tion (Cambrian),  are  mostly  mudstones  and  shales  but  the  highest  of  these  pass  upwards, 
apparently  without  a  break,  into  a  group  of  sandstones.  The  Bedinan  Formation  represents 
only  a  part  of  the  Caradoc  Series  and  there  is  an  abundant  shelly  fauna  which  includes  the 
following  trilobites  :  Ampyx  nitidus  sp.  nov.,  Marrolithoides  orthogonius  sp.  nov.,  M.  laticirrus 
sp.  nov.,  Cryptolithus?  inferus  sp.  nov.,  C?  bedinanensis  sp.  nov.,  Dionide  formosa  anatolica 
subsp.  nov.,  Dalmanitina  proaeva  (Emmrich),  Kloucekia  phillipsii  euroa  subsp.  nov.,  Neseuretus 
{Neseuretinus)  turcicus  subgen.  et  sp.  nov.,  Brongniartella  levis  sp.  nov.,  Plalycoryphe?  sp., 
Colpocoryphe  sp.,  Selenopeltis  inermis  angusticeps  subsp.  nov.  The  trilobites,  together  with  the 
less  common  brachiopods,  exhibit  marked  Bohemian/Tethyan  affinities  and  a  tentative  correla- 
tion with  the  Cernin  and  Chlustina  Beds  of  Czechoslovakia  is  suggested.  The  graptolite 
evidence,  though  fragmentary,  probably  indicates  the  multidens  or  clingani  Zone  and  is  broadly 
in  keeping  with  that  of  the  shelly  faunas. 

geol.  15,  2.  10 


84 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 
I.  INTRODUCTION  AND  ACKNOWLEDGMENTS 


Although  large  outcrops  of  Lower  Palaeozoic  rocks  occur  in  south-eastern  Turkey, 
between  the  Tigris  and  Euphrates  valleys,  relatively  little  published  information  is 
available.  The  best-known  exposures  lie  along  an  elongated  belt,  up  to  almost 
3  km.  wide,  running  east-south-east  from  a  point  2  km.  south-east  of  Derik  towards 
Mardin,  about  20  km.  north-west  of  the  Syrian  frontier  (for  place  names  see  Text- 
fig,  i).  Tolun  &  Ternek  (1952)  gave  a  short  description  and  small-scale  maps  of  the 
Cambrian  outcrops  near  Derik,  and  the  highest  part  of  the  Cambrian  succession  as 
shown  by  them  is  known  now  to  be  Ordovician  in  age.  Later  Tolun  (1960  :  236) 
noted  the  occurrence  of  Silurian  rocks  (sensu  lato  including  Ordovician)  in  south- 
eastern Turkey  and  mentioned  a  succession  of  900  m.  of  marly  and  sandy  beds  with 
brachiopods,  graptolites  and  trilobites  underlying  Cretaceous  limestones  at  Bedinan. 
He  remarked  also  on  the  similarity  of  the  Bedinan  Ordovician  rocks  to  others  found 
in  bore-holes  in  northern  Syria.  The  most  important  work  dealing  with  this  region 
is  that  of  Kellogg  (i960)  who  mapped  a  large  area  west  of  Mardin  and  gave  detailed 
sections  through  all  the  stratigraphical  subdivisions  present,  including  the  Cambrian 
and  Ordovician  rocks.  He  gave  no  comprehensive  faunal  lists  but  described  the 
Bedinan  Formation  as  containing,  especially,  "  Cryptolithus  "  and  "  Sowerbyella- 
like  brachiopods  ",  which  were  held  to  indicate  a  Middle  Ordovician  age.  Unfor- 
tunately Kellogg's  report  remains  unpublished,  but  reference  will  be  made  to  his 
work  from  time  to  time  in  this  paper.  A  chart  of  the  rock  succession  in  south- 
eastern Turkey,  from  Pre-Cambrian  to  Quaternary,  was  compiled  by  Gemot  Schmidt 


Fig.  1.     Sketch-map  of  south-eastern  Turkey  showing  location  of  place-names  mentioned 
in  text. 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  85 

in  1964  and  revised  in  1965.     Again,  this  is  a  publication  that  I  have  been  unable 
to  find  generally  available,  though  some  of  its  conclusions  are  noted  later. 

During  the  Spring  of  1965  my  wife  and  I  carried  out  field-work  in  the  region 
between  Derik  and  Mardin.  Both  the  Cambrian  and  Ordovician  successions  were 
examined  ;  the  latter  are  now  described  and  it  is  proposed  to  describe  the  Cambrian 
faunas  at  a  later  date.  Our  work  was  greatly  facilitated  by  the  kindness  of 
numerous  Turkish  friends.  The  Director,  Dr.  Sadrettin  Alpan,  and  other  members 
of  the  Maden  Tetkik  ve  Arama  Enstitiisu  (M.T.A.),  Ankara,  generously  placed  the 
facilities  of  their  organization  at  our  disposal  and  we  are  particularly  indebted  to 
Dr.  and  Mrs.  Kiragh,  as  well  as  to  Mr.  Giinal  Aygiin  who  helped  us  in  the  field. 
In  the  Derik  region  we  received  much  help  from  Dr.  Karakoyunlu  and  members  of 
the  M.T.A.  base-camp  at  the  Mazidag,  whilst  the  work  of  Mr.  Abdurrahman  Tung  as 
guide  and  collector  was  invaluable.  While  working  on  comparative  material  in 
Czechoslovakia  we  received  kindly  assistance  from  Dr.  Radvan  Horny  and  Dr. 
Ladislav  Marek.  The  graptolites  we  collected  were  examined  by  Dr.  Isles  Strachan. 
Finally,  Prof.  H.  B.  Whittington  read  the  manuscript  and  made  suggestions  for  its 
improvement.  All  specimens  described  in  the  present  work  are  in  the  collections  of 
the  Department  of  Palaeontology,  British  Museum  (Natural  History). 

II.  THE  SUCCESSION  AT  BEDINAN 

The  village  of  Bedinan  (sometimes  written  as  Badinan  or  Bahdinan)  lies  in  a 
valley  about  20  km.  east-south-east  of  Derik.  The  Ordovician  rocks  there  form  an 
inlier  about  6  km.  long  and  up  to  3  km.  broad  elongated  in  an  east-west  direction. 
The  inlier  is  bounded  to  the  north  by  a  prominent  scarp  and  plateau  of  unconformable 
Cretaceous  limestones,  with  a  gentle  northerly  dip,  and  to  the  south  by  a  large 
east-west  dislocation,  the  Mardin  Fault,  which  delimits  massive  Tertiary  limestones 
having  a  variable  southerly  dip  and  sometimes  slightly  overturned.  For  the  most 
part  the  Ordovician  rocks  comprise  mudstones  and  shales  with  some  intercalations  of 
calcareous  siltstone  ;  the  latter  sometimes  exhibit  cone-in-cone  structure  and  are 
more  resistant  to  erosion  than  the  adjacent  mudstones.  Higher  in  the  succession 
the  beds  become  more  arenaceous  and  pass  upwards  conformably  into  a  series  of 
current-bedded  sandstones,  the  age  of  which  is  discussed  later. 

The  principal  section  of  the  Bedinan  Formation  occurs  to  the  west  of  the  village. 
The  succession  there  is  almost  totally  argillaceous,  broken  only  by  occasional, 
sometimes  concretionary,  bands  of  siltstone.  The  strata  have  a  slightly  variable 
dip  of  rather  more  or  less  than  400  to  the  north-north-east,  and  by  means  of  a  traverse 
in  this  direction  it  was  possible  to  sample  the  rocks  in  ascending  order.  The  location 
and  horizons  of  the  principal  fossiliferous  localities  are  shown  in  Text-figs.  2  and  3. 
The  lowest  beds  presumably  occurring  here  form  a  low-lying  area  immediately 
adjacent  to  hills  of  more  resistant  Tertiary  limestones,  and  could  not  be  examined 
owing  to  the  cover  of  Alluvium.  Consequently  it  was  impossible  to  confirm  the 
suggestion,  made  elsewhere  in  this  paper,  that  these  lowest  strata  may  probably  be 
correlated,  at  least  in  part,  with  the  relatively  restricted  Ordovician  succession  of 
the  Sosink  district.  However,  the  lowest  beds  sampled  (at  localities  B.i  and  B.2) 
yielded   Neseuretus    (Neseuretinus)    turcicus    sp.    nov.,    Colpocoryphe    sp.,    Ampyx 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 


Fig.  2.  Sketch-map  showing  principal  fossil  localities  in  the  Ordovician  rocks  (horizontal 
shading)  near  Bedinan.  Diagonal  shading  denotes  Tertiary  strata,  whilst  the  remaining 
areas  are  covered  by  Alluvium. 

nitidus  sp.  nov.,  Cryptolithusl  and  Selenopeltis  inermis  angusticeps  ssp.  nov.  Three 
of  these  genera  are  known  from  near  Sosink  but  were  not  recorded  from  the  succeeding 
strata  near  Bedinan.  Dalmanitina  and  Kloucekia  did  not  prove  suitable  for  attempts 
to  subdivide  the  Bedinan  Formation,  and  were  found  to  persist,  virtually  unchanged, 
throughout. 

Perhaps  the  least  fossiliferous  group  of  strata  encountered  in  the  traverse  were 
those  in  the  middle  part  of  the  section,  including  localities  B .  5-B .  9.     Nevertheless, 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 


87 


sufficient  material  was  collected  to  show  that  the  fauna  differs  in  no  way  from  that  of 
the  higher  beds.  The  strata  in  this  and  lower  parts  of  the  succession  were  disturbed 
by  the  intrusion  of  a  dyke  alleged  by  Kellogg  (i960)  to  be  of  Quaternary  age.  The 
adjacent  shales  are  often  broken  and  collecting  there  is  difficult,  but  the  degree  of 
metamorphism  is  not  high  and  at  one  point,  near  locality  B  .9,  relatively  undistorted 
fossils  were  obtained  within  a  few  centimetres  of  the  contact. 


ALLUVIUM 


Z 

o 

I— 

< 

O 


z 
< 

z 

Q 

LU 

CO 


TERTIARY 


Fig.  3.     Diagrammatic  vertical  section  of  the  strata  just  west  of  Bedinan,  showing  horizons 
of  principal  Ordovician  fossil  localities.     Shading  as  in  Text-fig.  2, 


88  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

The  highest  group  of  strata  exposed  in  this  section  crop  out  along  a  hill-slope 
capped  by  superficial  deposits.  The  beds  mostly  comprise  grey-green  shales  which 
weather  yellowish-brown  and  crumble  easily.  They  are  sometimes  highly  fossili- 
ferous,  containing  especially  trinucleid  trilobites  (Cryptolithus?) ,  Dalmanitina  and 
Kloucekia  ;  the  specimens  are  often  compressed,  but  occasional  micaceous  siltstone 
bands  may  yield  less  distorted  material.  Brachiopods,  including  the  genera 
Aegiromena  and  Svobodaina,  of  Bohemian  affinities,  were  found  at  several  localities 
but  proved  particularly  abundant  at  B.18-B.21. 

According  to  Kellogg,  still  higher  strata  near  Bedinan  may  be  examined  in  the 
area  to  the  south-east  of  the  village,  and  a  traverse  was  made  in  that  direction. 
His  measured  section  shows  the  shale  succession  passing  upwards  into  a  thick  series 
of  sandstones,  the  age  of  which  he  put  at  "  Middle?  or  Upper?  Ordovician  ",  these 
being  followed  in  turn  by  unfossiliferous  sandstones  and  shales  of  the  Dadas  Forma- 
tion, of  "  Lower?  Silurian  "  age.  More  recently  Schmidt  (1965)  has  assigned  the 
Bedinan  Formation  to  both  the  Ordovician  and  the  Silurian,  followed  (though  the 
nature  of  the  contact  is  questionable)  by  the  Dadas  Formation,  of  alleged  Devonian 
age.  Whatever  the  relative  merits  of  these  two  schemes,  and  there  is  as  yet  no 
definite  faunal  evidence  to  support  either,  it  seems  clear  that  the  shales  and  mud- 
stones  forming  the  bulk  of  the  Bedinan  Formation  pass  upwards  conformably  into  a 
series  of  massive  and  flaggy,  often  current-bedded  sandstones.  The  higher,  more 
massive,  arenaceous  strata  yielded  no  more  than  occasional  indeterminate  fragments 
of  inarticulate  brachiopods,  but  in  the  lower  sandstones  I  was  able  to  find  occasional 
specimens  of  Dalmanitina  and  Kloucekia,  apparently  identical  with  the  forms 
occurring  so  abundantly  in  the  argillaceous  strata  near  Bedinan.  As  discussed 
elsewhere  in  this  paper,  the  faunas  of  the  argillaceous  Bedinan  Formation  suggest 
an  horizon  in  the  upper  half,  though  not  the  highest  part,  of  the  Caradoc  Series. 
Consequently  there  is  no  necessity  to  regard  at  least  the  lower  portion  of  the 
succeeding  sandstone  sequence  as  being  any  later  than  Caradoc  in  age. 

Faunal  List  (for  localities  see  Text-fig.  2) 

Ampyx  nitidus  sp.  nov.  B.2. 

Brongniartella  levis  sp.  nov.  B.18. 

Brongniartella  sp.  B.13,  19,  20. 

Cheirurid  gen.  et  sp.  indet.  B.12. 

Colpocoryphe  sp.  B.i,  2. 

Dalmanitina  proaeva  proaeva  (Emmrich)  B.1-3,  6,  7,  10,  n,  13-15,  17-22. 

Kloucekia  phillipsii  (Barrande)  euroa  subsp.  nov.  B.2-4,  6-8,  11-21. 

Cryptolithus?  inferus  sp.  nov.  B.i. 

Cryptolithus?  cf.  inferus  sp.  nov.  B.2,  3. 

Cryptolithus?  bedinanensis  sp.  nov.  B.6,  8-16,  20,  22. 

Marrolithoides  laticirrus  sp.  nov.  B.3,  4. 

Marrolithoides  sp.  B.18?  B.21. 

Neseuretus  (Neseuretinus)  turcicus  subgen.  et  sp.  nov.  B.2. 

Platycoryphe  ?  sp.  B.20. 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  89 

Selenopeltis  inermis  (Beyrich)  angusticeps  subsp.  nov.  B.i. 

Ostracoda  indet.  B.8,  17,  22. 

Aegiromena  sp.  B.1-3,  6,  8,  9. 

Lingula  sp.  B.7. 

Schizomania  sp.  B.i. 

Svobodaina  sp.  B.2,  6,  8,  10-12,  17-21. 

Trematis?  sp.  B.13. 

Ribeiria  sp.  B.3. 

Miscellaneous,  poorly-preserved  bivalves  B.1-3,  6-8,  16. 

Sinuites  (s.l.)  sp.  B.3. 

Hyolithids  indet.  B.io. 

Lepidocoleus  sp. 

Plumulites  sp.  B.io,  12,  13,  16,  22. 

Crinoid  fragments,  B.13,  20. 

Climacograptus  sp.  B.13. 

Diplograptus  sp.  B.8. 


III.  THE  SUCCESSION  NEAR  SOSINK 

Although  the  Bedinan  Formation  forms  an  elongated  E-W  outcrop  some  5  km. 
by  1  km.  just  east  of  the  village  of  Sosink,  the  strata  are  not  well  exposed.  Much 
of  the  ground  is  covered  by  superficial  deposits  derived  from  the  adjacent  high 
ground  to  the  east  where  Cretaceous  limestones,  dipping  south,  form  a  plateau-like 
unconformable  layer,  and  the  best  exposures  are  limited  to  a  section  running 
N.N.W.-S.S.E.  along  the  dip,  in  the  vicinity  of  the  small  stone  building  known  as 
Ziyaret  (Text-fig.  4).  The  rocks,  like  those  of  the  Bedinan  district,  consist  essentially 
of  grey-green  mudstones  and  shales  with  occasional,  harder  bands  of  siltstone.  The 
beds  dip  just  east  of  south  at  an  average  angle  of  about  380  though  with  slight  varia- 
tions, and  the  estimated  thickness,  calculated  on  the  basis  of  a  section  through 
Ziyaret,  is  of  the  order  of  440  m.  In  general  the  rocks  are  poorly  fossiliferous, 
though  specimens  may  be  locally  abundant,  occurring  in  thin  bands.  The  mud- 
stones  and  shales  are  deeply  weathered  and  crumble  easily  whilst  the  fossils  are 
almost  invariably  preserved  as  orange,  limonitic,  internal  and  external  moulds. 

The  northern  boundary  of  the  Ordovician  outcrop  here  is  a  fault,  separating  the 
beds  from  Cambrian  sandstones  which  form  a  conspicuous  feature  immediately  to 
the  north.  The  fault  has  an  indeterminate,  small  downthrow  to  the  south  and  is 
thought  to  mask  the  unconformable  base  of  the  Bedinan  Formation.  The  lowest 
Ordovician  beds  exposed  are  not  far  from  the  fault-line  and  proved  only  sparingly 
fossiliferous,  locality  A.i  yielding  a  single  specimen  of  Colpocoryphe  and  a  few 
poorly-preserved  brachiopods.  Much  of  the  Ordovician  succession  exposed  in  the 
hill-slopes  to  the  south-east  of  Ziyaret  proved  almost  barren,  and  only  a  small 
number  of  specimens  was  obtained  from  locality  A. 2  though  these  did  include 
Lasiograptus  sp.  and  the  holotype  cranidium  of  Selenopeltis  inermis  angusticeps 
subsp.  nov.  in  addition  to  the  more  commonplace  fauna  of  Dalmanitina,  Aegiromena 
and  bellerophontid  gastropods  [Sinuites  s.l.). 


go 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 


The  most  prolific  faunas  of  this  area  were  collected  from  a  number  of  localities  in  a 
small  N.E.-S.W.  valley  excavated  in  shales  some  180  m.  north-west  of  the  estimated 
position  of  the  Cretaceous/Ordovician  unconformity.  In  general  the  fossils  were 
found  in  restricted  horizons,  no  more  than  a  few  cm.  thick,  which  could  not  be 
traced  with  certainty  for  more  than  a  few  metres.  Trinucleid  trilobites  (Marro- 
lithoides)  formed  easily  the  most  abundant  constituent  of  the  fauna,  though 
Dalmanitina  and  Kloucekia  were  not  uncommon.     The  presence  of  Dionide  formosa 


a  l  L^uyiuMn.ui^ijjA-; :  \ > 


CRETACEOUS^q 

i ;  i .  i .  i ;  i ;  i ;  i  •  • 


r~r,  i    i  ,  i 


Fig.  4.  Principal  fossil  localities  in  the  Ordovician  rocks  near  Sosink,  8  kilometres  south- 
south-east  of  Derik.  Shading  as  in  Text-fig.  2,  with  addition  of  outcrops  of  Cambrian 
(dotted)  and  Cretaceous  (brick  pattern)  rocks.     Geological  boundaries  after  Kellogg  i960. 


anatolica  at  three  localities  was  of  particular  interest  ;  the  species  was  found  within  a 
thickness  of  only  a  few  metres  of  shales,  and  the  genus  is  not  yet  known  elsewhere  in 
south-eastern  Turkey.  Brachiopods  were  relatively  uncommon  in  the  sections  near 
Ziyaret,  and  comprised  only  Aegiromena,  no  doubt  the  "  Sowerbyella-like  brachio- 
pods "  of  Kellogg's  account.  Very  few  representatives  of  the  more  abundant 
brachiopod  faunas  (including  Svobodaina)  of  the  Bedinan  district  were  found, 
probably  owing  to  the  stratigraphically  lower  level  of  the  Ziyaret  strata.  The 
highest  Ordovician  strata  in  this  area  could  not  be  examined  owing  to  the  cover 
of  Recent  superficial  deposits,  which  also  obscures  the  junction  of  the  Ordovician 
and  Cretaceous  rocks. 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  91 

Faunal  List  (for  localities  see  Text-fig.  4) 

Asaphid  gen.  et  sp.  indet.  A. 6. 

Colpocoryphe  sp.  A. 3. 

Dalmanitina  proaeva  proaeva  (Emmrich)  A.  1-6. 

Dionide  formosa  (Barrande)  anatolica  subsp.  nov.  A. 3,  5,  7. 

Marrolithoides  orthogonius  sp.  nov.  A. 3-6. 

INeseuretus  (Neseuretinus)  turcicus  subgen.  et  sp.  nov.  A. 3. 

Selenopeltis  inermis  (Beyrich)  angusticeps  subsp.  nov.  A. 2,  3. 

Aegiromena  sp.  A. 2,  3,  4,  5,  7. 

Svobodaina  sp.  A. 3. 

Sinuites  (s.l.)  sp.  A. 2,  6. 

Redonia  sp.  A. 3. 

Ribeiria  sp.  A. 3. 

Bivalve  indet.  A. 3. 

Miscellaneous  Ostracoda  A. 3,  6. 

Hyolithids  indet.  A. 3. 

Plumulites  sp.  A. 3,  4,  5,  6,  7. 

Climacograptus  sp.  A. 3. 

Lasiograptus  sp.  A. 2. 

Orthocone  cephalopod  indet.  A. 3. 

IV.  AGE  AND  RELATIONSHIPS  OF  THE  BEDINAN  FAUNAS 

Perhaps  the  most  obvious  feature  of  the  Beninan  shelly  faunas  is  their  overall 
resemblance  to  those  of  the  Bohemian  region,  even  though  the  relevant  Turkish 
forms  are,  for  the  most  part,  at  least  subspecifically  distinct.  The  Bohemian  trilobite 
species  identical  with,  or  most  closely  related  to,  those  of  the  Bedinan  Formation 
are  as  follows  :  Dalmaniiina  proaeva  proaeva  (Emmrich),  Kloucekia  phillipsii 
(Barrande),  Dionide  formosa  (Barrande)  and  Selenopeltis  inermis  inermis  (Beyrich) 
[=  S.  buchi  (Barrande)].  From  accounts  of  the  Bohemian  faunas  published  by 
HavlFeck  et  al.  (1958)  and  Snajdr  (1956)  the  ranges  of  these  species  are  as  follows  : 
D.  proaeva  proaeva,  Cernin  &  Chlustina  Beds;  Dionide  formosa,  Cernin  Beds.  The 
lists  of  Havlicek  et  al.  show  that  Kloucekia  phillipsii  occurs  only  in  the  Chlustina 
Beds,  its  type  horizon,  but  Snajdr  records  it  from  the  Drabov  Beds  (basal  Caradoc 
Series)  to  the  Chlustina  Beds.  Selenopeltis  inermis  has  an  extended  vertical  range 
and  is  alleged  to  occur  as  low  as  the  Dobrotiva  Beds  (Llandeilo  Series)  and  as  high 
as  the  Bohdalec  Beds  (topmost  Caradoc).  The  species  has  been  recorded  (as 
S.  buchi)  by  Seilacher  (1963)  from  the  Sinat  Shales,  of  unspecified  Ordovician  age, 
in  northern  Iraq,  not  far  to  the  east  of  the  Bedinan  district. 

In  the  Bedinan  Formation  Dalmanitina  proaeva  proaeva  and  Kloucekia  phillipsii 
euroa  occur  throughout  most  of  the  fossiliferous  sequence,  an  association  suggesting 
that  a  tentative  correlation  with  the  combined  Cernin  Beds  and  Chlustina  Beds  is 
not  unreasonable.  Dionide  formosa  anatolica  has  been  found  only  in  the  Sosink 
district,  in  the  lower  part  of  the  Bedinan  Formation,  and  this  geological  horizon  may 
not  be  far  removed  from  that  of  D.  formosa  in  the  Cernin  Beds. 


92  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

The  evidence  of  the  remaining  trilobites  is  inconclusive,  though  the  species  present 
may  be  potentially  useful  for  correlation  when  other  Tethyan  faunas  are  better 
documented.  M arrolithoides  is  a  Llandeilo/Caradoc  genus  in  north-western  France 
and  the  Anglo-Welsh  area,  where  Cryptolithus  is  found  in  the  Llanvirn  and 
Llandeilo  Series,  whilst  Colpocoryphe  is  a  fairly  common  constituent  of  Mediterranean 
faunas  ranging  in  age  from  Arenig  to  Caradoc  Series.  Brongniartella  is,  above  all,  a 
genus  of  the  Caradoc  Series  in  the  Anglo-Welsh  area,  most  of  its  other  occurrences  in 
Europe  and  Scandinavia  being  as  a  single,  widespread  species,  B.  platynota  (Dalman), 
in  the  Ashgill  Series.  Its  appearance  in  the  Turkish  Caradoc  is  rather  unexpected 
but  suggests  an  easterly  migration  along  the  Tethys  during  Caradoc  times,  though  its 
subsequent  migration  and  development  are  far  from  clear.  The  new  subgenus 
Neseuretus  (Neseuretinus)  is  of  particular  interest  as  it  may  provide  a  link  between 
European  and  Asian  faunas  of  generally  similar  age.  The  horizon  of  N.  (Neseuretinus) 
birmanicus  (Reed)  in  Burma  and  China  is  obscure  in  terms  of  modern  stratigraphy, 
but  it  forms  part  of  Whittington's  (1966  :  723)  Caradoc  Encrinurella  fauna.  As 
pointed  out  elsewhere  (Dean  1967)  the  age  of  the  Encrinurella  fauna  may  vary 
within  the  Caradoc  Series,  and  is  apparently  greatest  in  Australasia. 

The  number  of  graptolites  found  during  the  present  field-work  was  disappointingly 
small  and  the  specimens  are  poorly  preserved,  but  Dr.  Strachan  has  kindly  examined 
them  and  supplied  notes  on  their  horizon.  They  include  Diplograptus  (s.s.)  sp. 
and  Climacograptus  sp.  near  Sosink.  Although  little  precise  information  can  be 
given  regarding  their  zonal  position,  Dr.  Strachan  considers  all  the  specimens  to  be 
of  Caradoc  age.  '  They  are  not  pve-gracilis  [Zone],  could  well  be  multidens-clingani 
[Zone],  and  are  unlikely  to  be  linearis  [Zone]  or  later  "  (personal  communication). 
Such  an  assessment  accords  well  with  the  evidence  of  the  trilobites,  and  supports  a 
broad  correlation  with  the  Cernin  and  Chlustina  Beds,  sub-divisions  which,  in 
Bohemia,  are  overlain  by  the  Bohdalec  Beds,  regarded  as  the  topmost  part  of  the 
Caradoc  Series  (Boucek  1937  :  454). 

All  the  other  animal  groups  represented  in  the  Bedinan  Formation  are  in  a  minority 
by  comparison  with  the  trilobites.  Brachiopods  may  be  locally  abundant,  partic- 
ularly in  the  higher  strata  west  of  Bedinan,  but  are  often  poorly-preserved.  They 
include,  inter  al.,  the  genera  Aegiromena  and  Svobodaina  (Havlicek  1950  :  38,  109), 
indicating  once  more  a  close  relationship  with  the  Caradoc  Series  of  Bohemia.  The 
remainder  of  the  fauna  consists  mainly  of  molluscs,  represented  by  the  almost 
ubiquitous  Redonia  and  nuculids,  together  with  gastropods,  usually  smooth  forms  of 
bellerophontid  type  generally  resembling  Sinuites.  The  alleged  phyllopod  genus 
Ribeiria  was  found  uncommonly  near  both  Bedinan  and  Sosink.  Hyolithids  occur 
in  small  numbers  but  are  usually  incomplete  and  poorly-preserved,  whilst  plates  of 
the  machaeridian  genera  Lepidocoleus  and  Plumulites  are  not  uncommon.  All  these 
groups  form  an  assemblage  broadly  similar  to  many  others  in  the  Tethyan  region, 
extending  as  far  west  as  Portugal,  Normandy,  Wales  and,  probably,  Florida.  Such 
assemblages  range  in  age  from  Arenig  to  Caradoc  Series  and  many  of  the  constituents, 
apart  from  the  more  obvious,  diagnostic  forms,  exhibit  relatively  little  morphological 
change,  although  a  modern  revision  of  the  molluscs  is  still  awaited. 


ORDO  V ICIAN  TRILOBITE  FAUN  A  OF  S.E.  TURKEY  93 

V.  SYSTEMATIC  DESCRIPTIONS 

Family  RAPHIOPHORIDAE  Angelin,  1854 

Genus  AMPYX  Dalman,  1828 

Ampyx  nitidus  sp.  nov. 
(PL  5,  figs.  5,  8-11) 

Diagnosis.  Ampyx  with  broad,  pear-shaped  glabella,  its  greatest  breadth 
measured  one-third  of  distance  from  base  of  frontal  spine  to  occipital  furrow.  Only 
traces  of  glabellar  lobes.  Triangular  fixigenae  relatively  short,  about  one  and  a 
half  times  as  broad  as  long.  Pair  of  large  pits  near  outer  ends  of  posterior 
border  furrow.  Pygidium  short,  with  median  length  about  one-third  frontal 
breadth.  Axis  poorly  defined  with  at  least  two  small  axial  rings.  Side  lobes  have 
one  pair  deep,  straight,  pleural  furrows. 

Holotype.     It.1181  (PI.  5,  fig.  8). 

Paratypes.  It.1180  (PL  5,  fig.  10)  ;  It.  1207  (PL  5,  fig.  5)  ;  It.  1208  (PL  5, 
fig.  9)  ;   It.1209  (PL  5,  fig.  11). 

Locality  and  horizon.  The  species  was  found  at  only  one  locality,  B.2,  south- 
west of  Bedinan,  in  the  lower  part  of  the  Bedinan  Formation  there.  It  was 
accompanied  by  Cryptolithust  inferus,  Neseuretus  (Neseuretinus)  turcicus  and 
inarticulate  brachiopods  (listed  as  Schizomania  sp.). 

Description.  The  cranidium,  excluding  the  frontal  spine  which  has  not  been 
found  preserved,  is  subtriangular  in  plan  with  median  length  about  half  of  the  basal 
breadth.  The  glabella  is  subtrapezoidal  in  outline,  bluntly  pointed  frontally,  with 
the  sides  diverging  forwards  at  about  500.  The  maximum  breadth  is  slightly  more 
than  three-quarters  of  the  median  length  of  the  glabella  (excluding  frontal  spine 
and  occipital  ring) ,  and  is  measured  across  the  intersection  of  the  sides  and  the  front 
of  the  fixigenae.  About  one-third  of  the  glabella  lies  in  front  of  the  line  of  maximum 
breadth.  Although  there  has  been  some  dorsal  compression,  there  can  be  no  doubt 
that  the  original  glabellar  outline  expanded  forwards  markedly.  The  glabella 
stands  a  little  higher  than  the  fixigenae  and  is  separated  from  them  by  broad,  shallow, 
slightly  curved  axial  furrows  which  become  deeper  frontally  and  curve  inwards, 
though  the  region  of  the  anterior  border  is  not  preserved.  The  glabella  carries  only 
poorly-defined  traces  of  lobation.  A  transverse,  basal  segment,  representing  the  ip 
glabellar  lobes,  is  delimited  by  a  pair  of  shallow  ip  glabellar  furrows  which  run 
inwards  and  slightly  forwards  to  join  medially,  where  they  become  almost  obsolete. 
Immediately  in  front  of  the  basal  glabellar  segment  are  traces  of  a  pair  of  semi- 
elliptical  2p  glabellar  lobes.  The  fixigenae  are  subtriangular  and  of  moderate  size, 
with  about  one-third  of  the  length  of  the  glabella  projecting  in  front  of  them.  The 
margins  are  straight  and  run  backwards  slightly  for  a  short  distance  from  the  axial 
furrows  as  far  as  the  facial  sutures.  Each  of  the  latter  meets  the  cephalic  margin 
at  an  acute  angle  and  from  there  pursues  a  gently  flexuous  course,  at  first  slightly 
concave  outwards  and  then  slightly  convex,  before  cutting  the  posterior  border 
immediately  outside  a  large,   slot-like  pit  in  the  posterior  border  furrow.     The 


94  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

occipital  ring  is  low,  moderately  broad  (sag.),  continuous  with,  but  projecting 
backwards  beyond,  the  transversely  straight,  ridgelike  posterior  border.  The 
occipital  furrow  is  mostly  shallow  but  deepens  a  little  abaxially  where  it  joins  the 
posterior  border  furrow.  The  latter  is  broad  (exsag.)  and  shallow  with  a  pair  of 
pits  sited  near  the  outer  ends  [see  above]. 

A  single  incomplete  hypostoma  was  found  (PI.  5,  fig.  10),  so  small  as  to  probably 
represent  an  immature  individual.  The  median  body  is  moderately  and  almost 
uniformly  convex,  bounded  posteriorly  and  posterolaterally  by  a  narrow,  slightly 
raised  border.  The  posterior  margin  is  almost  transversely  straight,  as  are  the 
posterior  portions  of  the  lateral  margins  which  converge  slightly  backwards  and 
meet  the  posterior  margin  at  rounded,  obtuse  angles. 

The  thorax  is  not  known. 

One  example  of  the  pygidium  was  found,  apparently  only  slightly  compressed. 
The  outline  is  transversely  subelliptical,  about  three  times  as  broad  as  long,  its 
frontal  margin  transversely  straight  except  medially,  where  a  small  articulating 
half-ring  projects  forwards  slightly.  The  axis  is  triangular  in  plan,  with  the  sides 
coverging  backwards  at  about  35 °.  It  is  not  strongly  denned,  stands  only  a  little 
higher  than  the  side-lobes,  and  reaches  the  posterior  margin  ;  two  axial  rings  are 
visible.  The  side-lobes  carry  one  pair  of  deep,  straight  pleural  furrows,  extending 
almost  to  the  margins  and  delimiting  a  pair  of  half-ribs  which  are  declined  antero- 
laterally  to  form  a  pair  of  small  facets.  The  impression  of  a  broad  doublure  is 
covered  with  fine  terrace-lines  which  run  subparallel  to  the  margin. 

Discussion.  Ampyx  nitidus  is  one  of  the  youngest  species  assigned  to  the  genus 
and  exhibits  marked  differences  from  early  forms  of  Ampyx.  The  type  species,  A. 
nasutus  Dalman  (see  Whittington  1950  :  554)  has  a  narrower,  less  divergent  glabellar 
outline,  with  a  smaller  proportion  of  the  glabella  projecting  in  front  of  the  fixigenae, 
which  are  also  relatively  longer.  The  Swedish  species,  which  is  of  Upper  Arenig 
age,  possesses  a  pair  of  pits  at  the  outer  ends  of  the  posterior  border  furrow,  as  does 
A.  nitidus.  This  feature  is  not  seen,  or  is  less  well  developed,  in  species  such  as 
Ampyx  linleyensis  Whittard  (1955  :  18),  from  the  Lower  Llanvirn  of  the  Shelve 
Inlier,  and  the  Shropshire  form  is  also  distinguished  by  its  broader  glabella  with 
smaller  frontal  projection,  as  well  as  by  the  well  differentiated  glabellar  and  alar 
lobes.  The  pygidium  of  the  Turkish  species  is  relatively  shorter  than  that  of 
Ampyx  nasutus  or  A.  linleyensis,  has  a  more  rounded  margin,  and  the  axis  is  less 
well  defined  and  segmented.  The  hypostoma  of  A.  nasutus  is  not  yet  known  but 
that  of  A.  linleyensis  has  a  pear-shaped  outline,  posterior  wings  and  lateral  notches 
not  seen  in  the  incomplete  specimen  attributed  here  to  A.  nitidus. 

Ampyx  virginiensis  Cooper  (1953  :  16),  from  the  Lower  Edinburg  Formation 
(early  Caradoc)  of  Virginia,  has  been  redescribed  in  detail  by  Whittington  (1959  :  465). 
Like  Ampyx  nitidus  it  has  a  relatively  short,  rounded  pygidium  but  differs  from  the 
Turkish  species  in  having  a  shorter  frontal  projection  of  the  glabella,  as  well  as 
more  distinct  glabellar  furrows  and  a  strong  development  of  alar  lobes.  As  far  as 
can  be  judged  the  hypostoma  of  A.  nitidus  appears  to  be  more  like  that  of  A. 
virginiensis  than  that  of  A.  linleyensis.  Ampyx  camurus  Raymond  (Whittington 
1959,  pi.  30,  figs.  15,  18,  19),  also  from  the  Edinburg  Formation,  has  a  cranidium 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  95 

generally  similar  to  that  of  A.  nitidus  but  the  glabella  is  slightly  narrower  and  the 
fixigenae  are  relatively  longer,  whilst  the  pygidium  has  a  more  triangular  outline. 
Ampyx  nitidus  bears  a  general  resemblance  to  A.  abnormalis  Yi  (1957  :  557,  pi.  5, 
figs.  3a-e),  also  of  Caradoc  age,  from  the  Yangtze-Gorge  District  of  China,  but  the 
Turkish  species  may  be  distinguished  by  its  slightly  longer  cephalon,  with  the  front 
of  the  glabella  extending  a  little  farther  in  front  of  the  fixigenae,  and  the  slightly 
shorter,  more  rounded  pygidium. 


Family  TRINUCLEIDAE  Hawle  &  Corda,  1847 

Subfamily  CRYPTOLITHINAE  Angelin,  1854 

Marrolithoides,  nowadays  accorded  generic  rank,  was  erected  by  Williams  (1948  : 
78)  as  a  subgenus  to  differentiate  Marrolithus-like  trilobites  in  which  the  lateral 
cephalic  margins  are  approximately  parallel,  the  arrangement  of  fringe-pits  is 
relatively  simple,  and  the  cephalic  fringe  is  uninflated  anterolaterally,  except  in 
gerontic  forms.  Since  then  Whittard  (1956  :  49,  63)  has  redefined  the  genus  and 
transferred  William's  species  M.  anomalis,  regarded  by  its  author  as  atypical  of 
Marrolithoides,  to  M arrolithus .  According  to  Whittard  the  criteria  for  distinguishing 
Marrolithoides  are  now  as  follows  :  (a)  the  cephalic  outline  is  subrectangular  ;  (b)  the 
fringe  is  not  distended  anterolaterally  and  there  is  no  abnormal  increase  in  pit- 
diameter  there  ;  (c)  auxiliary  pits  are  generally  present  in  Ex  ;  (d)  E1(  Ix  and, 
sometimes,  I2  are  continuous  frontally  ;  (e)  lx-l2  pseudogirder  is  almost  as  strongly 
developed  as  the  normal  E^-^  girder.  Specimens  from  the  Bedinan  Formation 
which  possess  these  features,  together  with  unmistakable  anterolateral  angulation 
of  the  cephalic  outline,  may  therefore  reasonably  be  assigned  to  Marrolithoides. 
Individuals  of  this  type  occur  in  the  succession  east  of  Sosink  and  also  at  locality 
B.3,  in  the  lowest  part  of  the  succession  exposed  near  Bedinan,  but  the  numerous 
trinucleids  collected  suggest  that  such  generic  limits  as  those  listed  above  may  be 
somewhat  arbitrary.  The  angular  cephala  of  B.3  are  associated  with  smaller 
individuals — apparently  immature  examples  of  the  same  species — in  which  the 
outline  is  rounded  anterolaterally.  Similarly-rounded  cephala,  but  of  relatively 
large  size  and  therefore  presumably  adult  forms,  occur  at  locality  B.i  as  well  as 
from  B.5  to  the  top  of  the  succession  near  Bedinan.  Trinucleid  cephala  possessing 
such  rounded  outlines  together  with  a  single  row  of  pits  external  to  the  girder  would 
normally  be  termed  Cryptolithus  sensu  stricto,  and  the  name  is  used  here,  with  some 
doubt,  for  most  of  the  Turkish  specimens  described.  The  latter,  in  general,  show  a 
greater  development  of  concentric  rows  of  pits  than  is  customary  for  such  forms  as 
the  type  species  C.  tesselatus  Green,  from  the  Caradoc  of  eastern  North  America, 
and  it  is  likely  that  the  affinities  of  the  Turkish  specimens  lie,  rather,  with  species 
in  Bohemia  as  well  as  those  described  by  Whittard  (1958  :  72-77)  from  the  Llanvirn 
and  Llandeilo  Series  of  the  Shelve  Inlier.  The  latter  group  of  species  was  said  by 
Whittard  to  be  distinct  from  those  in  North  America,  but  there  is  as  yet  no  evidence 
that  they  merit  generic  separation.  On  the  basis  of  the  above  criteria  the 
trinucleids  of  the  lowest  Bedinan  Formation  are  placed  in  Marrolithoides.     The 


96  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

remainder  are  attributed  questionably  to  Cryptolithus  and  there  is  a  small  strati 
graphical  overlap  of  the  two  genera  in  the  lower  strata  exposed  south-west  of  Bedinan. 
The  terminology  used  in  the  following  descriptions  is  that  of  Whittard  (1955  :  27) 
and  the  pit  counts  refer  to  half  the  cephalic  fringe. 

Genus  MARROLITHOIDES  Williams,  1948 

Marrolithoides  orthogonius  sp.  nov. 

(PL  1,  figs.  1-9) 

Diagnosis.  Marrolithoides  with  subrectangular  cephalic  outline  and  three 
concentric  rows  of  pits  [E1(  I]_2]  developed  frontally,  except  for  small  irregular 
group  near  sagittal  line.  E1(  I2  pits  of  similar  size  ;  I1  pits  slightly  larger.  I3_5 
present,  but  I6  represented  by  only  few  small  pits.  Average  pit-count  :  Ex  +  ex  26, 
Ix  +  ix  21,  I2  +  i2  21,  I3  +  i3  18,  I4  +  i4  17,  I5  +  i5  about  14  or  15  [for 
variation  see  description]. 

Holotype.     It.  1200  (PL  1,  fig.  6). 

Paratypes.  It. 747  (PL  i,  fig.  4)  ;  It. 749  (PL  1,  fig.  8)  ;  It. 760  ;  It. 762  (PL  1, 
fig.  7)  ;  It. 803  (PL  i,  fig.  2)  ;  It. 806  (PL  r,  fig.  5)  ;  It. 812  (PL  1,  fig.  1)  ;  It. 818 
(PL  1,  fig.  9);   It.8i9(Pl.i,fig.  3). 

Localities  and  horizons.  The  holotype  and  most  of  the  paratypes  are  from 
the  Bedinan  Formation  at  locality  A. 3  in  the  section  near  Ziyaret,  some  1300  m. 
east  of  Sosink.  This  locality  yielded  the  species  in  greatest  abundance,  some  60 
specimens,  whilst  more  than  20  were  obtained  from  the  same  section  at  locality  A .  6, 
whence  came  two  of  the  paratypes.  Localities  A. 4  and  A. 5  produced  only  a  few 
specimens  of  M.  orthogonius. 

Description.  The  cephalon  is  approximately  subrectangular  in  outline,  more 
than  twice  as  broad  as  long  (excluding  spines),  though  the  proportions  are  obviously 
affected  by  dorsal  compression,  so  that  the  cephalic  fringe  now  appears  flattened, 
in  contrast  to  its  original,  steep  declination  outwards.  The  frontal  margin,  gently 
convex  forwards  in  plan,  forms  a  broad  curve  and  meets  at  an  obtuse  angle  the 
lateral  margins,  which  are  almost  straight  and  converge  forwards  gently.  The 
degree  of  anterolateral  angulation  seems  to  be  genuinely  variable,  ranging  from  a 
broad  curve,  particularly  in  smaller  specimens,  to  a  relatively  sharp  angle.  In  one 
case  (PL  1,  fig.  5)  the  angle  is  marked  by  a  blunt  projection  developed  from  the 
margin  at  about  R  16.  The  glabella  is  about  twice  as  long  as  broad,  stands  high  above 
the  cheek-lobes,  and  narrows  backwards  slightly  to  a  shallow  occipital  furrow  which 
deepens  laterally  into  a  pair  of  apodemal  pits.  The  best-preserved  specimens  show 
a  small,  median  tubercle  which  is  slightly  more  conspicuous  on  the  internal  than  on 
the  external  mould.  The  glabella  extends  forwards  a  little  beyond  the  cheek-lobes, 
from  which  it  is  separated  by  deep,  straight,  axial  furrows,  so  that  the  cephalic 
fringe  narrows  (sag.)  frontally.  Each  axial  furrow  contains  a  prominent  hypostomal 
pit  sited  just  behind  the  pitted  fringe.  Some  specimens  show  faint,  lateral  indenta- 
tions of  the  glabella,  suggesting  almost  obsolete  glabellar  furrows,  but  in  most  the 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  97 

glabellar  margins  are  entire.  One  of  the  smallest  specimens  (Meraspis,  Degree 
unknown,  PL  1,  fig.  7)  shows,  in  addition  to  traces  of  eye-ridges,  a  conspicuous 
development  of  alar  lobes  ;  the  latter  are  less  obvious  on  both  an  example  of 
Meraspis,  Degree  4  (PL  1,  fig.  9)  and  slightly  larger  cephala  (PL  1,  figs.  1,  2),  and  are 
absent  from  the  presumed  adults.  A  deep  occipital  furrow  separates  the  glabella 
from  the  narrow  (sag.)  occipital  ring,  which  is  steeply  inclined  backwards,  produced 
upwards  and  back  to  form  a  thorn-like  occipital  spine  approximately  half  the  length 
of  the  glabella.  The  posterior  border  is  narrow  (exsag.)  and  transversely  almost 
straight  as  far  as  the  fulcra,  where  it  flexes  backwards  a  little  way,  becoming  flange- 
like and  indented  to  form  a  pair  of  articulating  sockets.  Beyond  the  fulcra  the 
posterior  margins  of  the  cephalic  fringe  are  almost  transversely  straight,  and  the 
backward  projections  of  the  fringe  (so-called  posterior  wings)  which  characterize 
certain  of  the  Bedinan  Formation  trinucleids,  and  are  sometimes  very  large,  are  here 
noticeably  absent.  The  small  Meraspis  figured  here  (PL  1,  fig.  7)  shows  the  genal 
angles  set  well  forwards,  in  front  of  the  line  of  the  posterior  border  furrow.  This 
feature  is  shared  with  other  cryptolithinids,  and  the  position  of  the  genal  angles 
moved  progressively  backwards  during  ontogeny.  Only  incomplete  examples  have 
been  found  showing  the  librigenal  spines  ;  these  are  directed  backwards  and  slightly 
outwards  from  the  genal  angles,  and  apparently  resemble  those  found  in  other 
members  of  the  Cryptolithinae.  Many  specimens  exhibit  a  conspicuous  reticulation 
of  the  cephalic  test,  but  in  others  it  is  less  well  developed  or  even,  occasionally,  almost 
absent.  It  is  not  yet  clear  whether  such  variation  is  original  or  due  to  vagaries 
of  preservation.  When  present  the  reticulation  is  confined  to  the  cheek-lobes  and 
the  axial  portion  of  the  glabella  (see  PL  1,  fig.  1). 

The  cephalic  fringe  is  narrow  (sag.)  in  front  of  the  glabella,  becomes  broader 
laterally,  attaining  its  maximum  opposite  the  antero-lateral  portions  of  the  cheek- 
lobes,  and  narrows  a  little  laterally,  finally  broadening  again  near  the  posterior 
border,  where  it  expands  around  the  posterolateral  extremities  of  the  cheek-lobes. 
Three  concentric  rows  of  pits,  Ex  and  I^,  are  almost  continuous  frontally,  except 
for  a  small,  irregular  area  near  the  sagittal  line  which  appears  to  be  of  specific 
importance.  At  this  point  small  cephala  (see  PL  1,  fig.  4)  exhibit  a  group  of  three 
pits  arranged  in  a  triangle  with  apex  directed  backwards.  The  anterior  two  pits 
form  part  of  E-^  whilst  the  hindmost  pit  may  probably  be  regarded  as  part  of  Iv 
though  sometimes  it  occupies  a  position  between  Ix  and  I2.  In  larger  cephala  (see 
PL  1,  fig.  8)  the  group  of  three  pits  is  replaced  by  one  of  four  pits,  three  of  them 
corresponding  to,  and  arranged  in-line  with,  E1(  and  the  centre  pit  of  the  three 
arranged  radially  with  the  fourth  pit  so  as  to  form  a  radial  row,  Ro,  coincident  with 
the  sagittal  line.  Additional  concentric  rows  of  pits  are  developed  as  follows  : 
I3fromR3  or  R4;  I4,  from  R6  or  R7  (occasionally  R5)  ;  I5  from  Rg  (less  commonly 
from  Rio  or  Rn)  ;  I6,  when  present,  is  developed  only  as  a  few  pits  in  the  area 
denoted  by  Rn  to  R15,  or  thereabouts.  The  pits  of  Ex  and  I2  are  of  similar  size 
and  show  almost  no  variation  in  size  over  the  whole  of  their  length.  The  pits  of  Ix 
are  slightly  bigger,  also  uniform  in  size,  and  a  low  ridge  is  developed  between  Ex 
and  I1(  particularly  anterolaterally.  Although  all  the  material  is  somewhat 
compressed  there  is  a  suggestion  that  the  corresponding  portions  of  I1  may  have 

GEOL.    15,  2.  II 


98  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

been  raised  slightly  above  the  adjacent  rows,  but  there  is  no  indication  of  distended 
pits  such  as  are  found  in  the  analogous  parts  of  Marrolithus.  The  pits  internal  to 
1 2  diminish  slowly  in  size  towards  the  cheek-lobes  and  are  often  set  in  shallow  radial 
sulci ;  the  latter  may  be  exaggerated  by  crushing,  which  may  also  overemphasize 
the  radial  extension  of  such  sulci  to  include,  apparently,  pits  of  Ix  and  I2  ;  The 
number  of  pits  along  the  posterior  border  of  the  fringe  is  generally  about  eight,  in 
addition  to  a  single,  larger,  apodemal  pit.  Apart  from  the  area  of  irregular  pitting 
near  the  sagittal  line  described  above,  the  arrangement  of  pits  is  remarkably  orderly, 
with  a  strong  radial  arrangement  persisting  from  R±  to  within  a  few  pits  of  the  genal 
angles.  Auxiliary  pits  occur  only  uncommonly,  though  an  occasional  specimen 
may  have  an  extra  pit  or  so  on  one  side  of  the  cephalon  and  not  on  the  other.  An 
example  is  shown  in  PI.  I,  fig.  6,  with  Ex  containing  two  intercalated  pits,  between 
R3  and  R4,  and  between  R8  and  Ro,  ;  these  occur  only  on  the  right  side  of  the 
cephalon.  The  number  of  pits  present  in  the  fringe  is  as  follows,  the  first  number 
indicates  the  number  of  pits  most  commonly  found,  the  second,  in  brackets,  shows 
the  range  of  variation  :  Ex  26  (23-28),  I1  21  (20-24,  rarely  25),  I2  21  (20-22,  rarely 
23),  I3  18  (17-19,  one  specimen  with  21),  I4  17  (15-18),  I5  approx.  14-15,  but  most 
material  not  suitably  preserved  ;  I6  either  not  developed,  or  present  only  as  a  few 
pits  from  a  point  varying  from  Ru  to  R15. 

The  ventral  side  of  the  cephalic  fringe  carries  an  Ej/Ij  girder  which  is  well  defined 
anteriorly,  less  so  laterally,  and  finally  becomes  more  pronounced  again  just  before 
attaining  the  genal  angle,  where  it  forms  a  well-marked  ridge  which  is  continued  along 
the  librigenal  spine.  Between  1^  and  I2,  and  between  I2  and  I3,  are  pseudogirders, 
each  successively  a  little  less  strongly  developed  than  the  last  but  nevertheless  well 
defined,  and  traces  of  additional  pseudogirders  occur  between  the  remaining  I  rows. 

The  thorax  is  known  only  from  poorly-preserved  material  of  characteristic 
cryptolithinid  form,  that  is  to  say  it  contains  six  segments,  the  first  of  them  macro- 
pleural.  Each  pleura  ends  in  a  blunt  point,  directed  posterolaterally,  and  carries 
a  broad  (exsag.),  straight,  shallow,  pleural  furrow  which  runs  gently  backwards 
abaxially  from  the  axial  furrow  almost  to  the  pleural  tip,  near  which  it  curves 
backwards  slightly  and  dies  out. 

The  pygidium  is  subtriangular  in  plan  with  the  transversely  straight  anterior 
margin  broken  only  by  the  articulating  half-ring.  The  lateral  margins,  defined  by  a 
small,  raised  ridge,  are  straight  and  widely  divergent  over  the  posterior  two-thirds  but 
then  curve  forwards  to  the  anterolateral  angles.  The  axis  stands  slightly  higher  than 
the  flattened  side-lobes  and  is  gently  rounded  in  cross-section.  The  anterior  two-fifths 
of  the  axis  carry  three  well-defined  axial  rings,  gently  curved  and  convex  forwards 
in  plan.  The  rings,  which  are  not  sharply  delimited  laterally,  are  separated  by  ring 
furrows  which  are  continuous  laterally  with  markedly  shallower  furrows  traversing 
the  side-lobes  and  running  gently  backwards  to  reach  the  marginal  rim.  The 
remainder  of  the  axis  has  traces  of  several  small,  poorly-defined  rings,  and  the  tip 
merges  into  the  marginal  rim,  here  less  sharply  defined.  The  side-lobes  carry  only 
traces  of  furrows  in  addition  to  those  continuous  with  the  first  three  ring  furrows. 
The  pygidium  of  Meraspis,  Degree  4  (PL  1,  fig.  9)  is  semielliptical  in  plan  and 
proportionately  shorter  than  that  of  the  adult  trilobite. 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  99 

Discussion.     For    convenience    the    species    of    Marrolithoides    are    discussed 
together  (see  p.  100). 


Marrolithoides  laticirrus  sp.  nov. 
(PL  2,  figs.  1,  3,  5,  9,  11,  13,  14) 

Diagnosis.  Marrolithoides  with  cephalic  outline  broadly  rounded  frontally, 
angular  anterolaterally.  Large  posterolateral  extensions  of  fringe.  Reticulation 
of  glabella  and  cheek  lobes  characteristic  in  both  small  and,  to  lesser  degree,  larger 
individuals.  Broad  cephalic  fringe  contains  four  continuous  concentric  rows  of  pits 
(E1;  Ij_3)  frontally  ;  remaining  rows  well  developed.  Pit  count  relatively  high, 
average  as  follows:  Ex 36-38,  Ij  27-30,  I2  26-30,  I3  29,  I4  up  to  26,  I5  21,  small 
development  of  I6.  Pygidium  with  about  nine  axial  rings  and  five  pairs  pleural 
ribs. 

Holotype.     It. 683  (PL  2,  fig.  5). 

Paratypes.  It. 690  (PL  2,  fig.  11)  ;  It. 706  (PL  2,  fig.  3)  ;  It. 707  (PL  2,  fig.  1)  ; 
It. 708  (PL  2,  fig.  13)  ;   It. 712  (PL  2,  fig.  14)  ;   It. 738  (PL  2,  fig.  9). 

Localities  and  horizons.  The  type  material  is  from  locality  B.3,  south-west 
of  Bedinan,  where  it  was  found  in  moderate  abundance  (sample  of  22  specimens) 
associated  with  Dalmanitina,  Kloucekia  and  a  single  example  of  Cryptolithus?  inferus. 
A  few  specimens  from  B  .4,  about  12  m.  higher  in  the  succession,  differ  from  the  type 
material  in  only  small  details  and  are  considered  to  fall  within  the  limits  of  variation 
for  the  species. 

Description.  The  cephalon  has  a  maximum  breadth  about  two  and  a  quarter 
times  the  median  length.  The  lateral  margins  are  straight  or  very  slightly  concave, 
and  typically  almost  parallel,  although  some  specimens  exhibit  a  slight  divergence 
or  convergence  which  may  have  been  exaggerated  by  crushing.  The  anterior 
margin  of  the  cephalon  is  arched  forwards,  moderately  in  smaller  cephala  but  more 
strongly  in  larger  specimens.  One  of  the  latter,  the  holotype  (PL  2,  fig.  5),  shows  the 
lateral  and  anterior  margins  meeting  at  obtuse  but  sharply  defined  angles  which  are 
in  line  with  the  highest  point  of  the  glabella,  about  one-third  of  the  distance  from 
the  front  of  the  glabella.  The  cephalic  fringe  is  notably  broad,  with  a  frontal 
breadth  (sag.)  [measured  on  dorsally  compressed  specimens]  equal  to  two-fifths  of 
the  glabellar  length.  Its  maximum  breadth  is  anterolaterally,  behind  which  it 
narrows  a  little,  just  in  front  of  the  line  of  the  posterior  border  furrow,  before  expand- 
ing again  towards  the  genal  angles,  which  are  set  well  back,  at  the  end  of  large  genal 
prolongations.  At  the  genal  angles  the  lower  lamella  of  the  fringe  is  produced 
backwards  to  form  a  pair  of  prismatic  librigenal  spines,  curving  gently  outwards  at 
first  and  then  inwards,  with  a  length  at  least  two  and  a  half  times  that  of  the  glabella. 
There  is  a  strong,  concentric  arrangement  of  fringe-pits,  usually  with  four  continuous 
rows  (E1;  Ij_3)  developed  in  front  of  the  glabella,  but  occasionally  only  three  rows 
(E1;  I±-2).  I4  may  be  developed  as  far  forwards  as  R3.  The  pits  of  I1  are  conspicu- 
ously the  largest  and  maintain  an  almost  uniform  size  to  the  genal  angle.     The 


ioo  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

pits  of  I2  are  slightly  smaller  whilst  those  of  Ex  and  E3  are  slightly  smaller  still. 
The  pits  of  subsequent  I  rows  diminish  steadily  in  size  inwards  from  the  girder. 
All  the  specimens  are  moderately  compressed  dorsally  but  most  show  a  development 
of  a  sharp  concentric  ridge  sited  anterolaterally  between  Ex  and  Ix,  and  extending 
from  about  R6  to  R.2  or  thereabouts.  Some  show  a  suggestion  of  a  smaller  Ix_2 
ridge  anterolaterally.  These  structures  suggest  that  the  uncrushed  cranidia  may 
originally  have  had  the  anterolateral  portions  of  Ix  raised  slightly  above  neighbouring 
rows.  The  number  of  pits  in  each  of  the  rows  Ix_4  is  fairly  uniform  but  the  number 
in  Ex  is  conspicuously  higher  and  many  of  the  pits  there  do  not  fit  easily  into  a  radial 
pattern.  The  region  of  the  fringe  internal  to  the  girder  exhibits  a  strong  radial 
arrangement,  comprising  from  seventeen  to  nineteen  radii  which  cover  an  area 
extending  to  within  a  few  pits  (generally  3  or  4)  of  the  posterior  border.  A  sample  of 
15  specimens  from  the  type  locality  yielded  the  following  counts  :  Ex  +  e1  36-38 
[in  small  cephala  29-34,  rarely  26]  ;  Ix  +  ix  26-30  [occasionally  23-25]  ;  I2  +  i2 
26-30  [occasionally  22-25]  ;  I3  +  i3  26-29  [occasionally  23-25]  ;  I4  +  i4  is  well 
developed  in  the  holotype,  with  27  pits  from  R3,  but  small  cephala  may  have  as  few 
as  18  pits  ;  I5  -f-  i5  10-19  [from  R8  in  a  large  cephalon,  from  R12  to  R14  in  a  small 
one]  ;  I6  only  a  trace,  perhaps  a  few  pits  from  about  R14-R16.  At  locality  B.4  a 
small  sample  of  three  cephala  was  obtained.  These  have  a  smaller  pit  count  (33-34) 
for  Ex  +  Gi  than  the  type  material  but  are  otherwise  similar  and  are  attributed  to 
the  same  species.  The  glabellar  outline  is  clavate,  expanded  frontally  where  the 
maximum  breadth  is  slightly  less  than  two-thirds  of  the  median  length.  Two 
pairs  of  lateral  depressions  represent  the  glabellar  furrows,  and  the  frontal  glabellar 
lobe  does  not  invade  the  cephalic  fringe  to  any  marked  degree.  The  cheek-lobes  are 
plump,  quadrant-shaped,  and  their  dorsal  surface,  like  that  of  the  axial  portion  of  the 
glabella,  is  covered  with  a  fine,  mesh-like  pattern  of  raised  ridges.  An  occasional 
specimen  shows  a  small,  median  tubercle  at  the  apex  of  the  glabella.  The  small 
occipital  ring  is  produced  backwards  and  upwards  to  form  a  small,  sharp  occipital 
spine.  On  the  ventral  surface  of  the  fringe  the  girder  is  moderately  developed, 
only  slightly  stronger  than  the  pseudogirders  IJI2  and  I2/I3. 

The  remainder  of  the  exoskeleton  is  generally  similar  to  that  found  in  other 
members  of  the  subfamily,  with  six  thoracic  segments  of  characteristic  type.  The 
subtriangular  pygidium  has  slightly  flexed,  steeply  declined  margins  surmounted 
by  a  thin,  dorsal  rim.  There  are  about  nine  small  axial  rings  separated  by  ring 
furrows,  the  anterior  members  of  which  cross  the  shallow  axial  furrows  and  are 
continuous  with  five  or  six  shallow  pleural  furrows. 

Discussion.  Marrolithoides  orthogonius  is  the  earliest-occurring  trinucleid 
species  yet  known  in  the  Bedinan  Formation,  and  was  found  only  in  the  upper  part 
of  the  section  east  of  Sosink.  The  not  inconsiderable  thickness  of  underlying  shales 
there  has  yet  to  yield  trinucleids,  but  the  beds  in  question  are  inadequately  exposed. 
M.  orthogonius  has  not  been  found  outside  the  Ziyaret  district  nor,  conversely,  have 
any  of  the  trinucleid  species  from  Bedinan  been  found  elsewhere.  This  may  be 
accounted  for  by  the  fact  that  the  Ziyaret  strata  are  probably  older  than  the  lowest 
seen  west  of  Bedinan,  and  if  any  overlap  of  the  sections  occurs  one  would  expect  it 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  101 

to  involve  the  strata  under  the  alluvial  cover  adjacent  to  the  Tertiary  limestones 
cropping-out  some  750  m.  south-west  of  Bedinan. 

Judging  from  Williams's  drawings  (1948,  pi.  6,  figs.  5,  7)  the  Welsh  forms  of 
Marrolithoides  would  appear  to  have  very  small  fringe  pits,  but  Whittard's  (1956, 
pi.  9,  fig.  3)  photograph  of  the  holotype  of  M.  simplex,  the  type  species,  shows  pitting 
of  characteristic  cryptolithinid  dimensions,  with  the  pits  of  Ix  slightly  larger  than  the 
rest,  a  feature  seen  also  in  the  Turkish  material.  The  Anglo-Welsh  species  of 
M arrolithoides  illustrated  by  Whittard  (1956)  possess  cheek-lobes  which  are  relatively 
much  larger  than  those  of  the  Turkish  species  ;  consequently  the  cephalic  fringes  of 
the  latter  appear  much  broader  although,  in  terms  of  concentric  rows  of  pits,  there 
is  not  much  difference.  M.  orthogonius  has  a  pit  count  not  markedly  different  from 
that  of  M.  arcuatus  Whittard,  of  lowest  Caradoc  age  at  Shelve.  However,  the 
Turkish  species  tends  to  have  more  pits  in  Ex  and  a  larger  development  of  I5,  whilst 
a  triangular  group  of  3  or  4  pits  (1  pit  in  Ix  and  2  or  3  pits  in  Ex)  forms  a  conspicuous 
feature  at  the  sagittal  line.  M arrolithoides  laticirrus  differs  markedly  from  all  other 
species  of  the  genus  in  its  large,  broad  fringe  with  four,  almost  five  concentric  rows 
of  pits  in  front  of  the  glabella  in  the  adult  trilobite.  Also  notable  is  the  large  number 
of  pits,  including  intercalated  ones,  in  Ex.  The  girder  of  M.  orthogonius  is  more 
strongly  developed  than  that  of  M.  laticirrus,  and  in  this  respect  the  latter  species 
has  a  ventral  aspect  more  reminiscent  of  the  species  of  Cryptolithus  ?  from  the  Bedinan 
Formation. 

The  French  species  Trinucleus  bureaui  Oehlert  (1895  :  300)  was  described  from  an 
unspecified  Ordovician  horizon  in  Brittany.  Whittard  (1956  :  54)  assigned  T. 
bureaui  to  Marrolithus  but  claimed  it  as  a  composite  species  and  drew  attention  to 
the  close  resemblance  of  some  of  Oehlert's  illustrations  to  M arrolithoides  simplex 
(Williams).  The  Breton  species  is  in  need  of  modern  revision  and,  as  described,  may 
well  include  material  attributable  to  both  Marrolithus  and  M arrolithoides.  Some  of 
the  specimens  figured  by  Oehlert  (e.g.  1895,  pi.  1,  figs.  1,  3)  generally  resemble  M. 
orthogonius  but  the  fringe  is  slightly  broader  anterolaterally,  apparently  the  result  of 
an  extra  concentric  row  of  pits  there,  and  lacks  the  characteristic  median,  triangular 
group  of  pits.  The  original  of  Oehlert's  pi.  1,  fig.  15  compares  with  a  paratype  of 
M.  orthogonius  (see  PI.  1,  fig.  5)  but  has  one  more  pit  row  and  greater  differentiation 
in  pit  size  than  the  Turkish  form. 

Marrolithoides  sp. 
(PI.  4,  fig-  10) 

In  general,  Marrolithoides  occurs  in  the  lower  part  of  the  Bedinan  Formation,  not 
only  near  Sosink  but  also  south-west  of  Bedinan,  where  it  overlaps  stratigraphically 
with  Cryptolithus  ?  inferus.  The  dominant  trinucleid  higher  in  the  Bedinan  Formation 
is  Cryptolithus}  bedinanensis  but  in  some  of  the  stratigraphically  highest  localities 
this  is  accompanied  by  uncommon  forms  which  match  best  with  Marrolithoides. 
The  largest  and  most  complete  of  these,  It. 881,  is  figured  here  from  locality  B.21. 
It  has  a  greater  number  of  concentric  pit  rows  than  M.  orthogonius,  and  although 
there  are  four  frontal  rows  as  in  M.  laticirrus  it  is  separated  from  that  species  by 


102  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

having  a  straighter  anterior  margin  and  fewer,  slightly  larger  pits  in  E1;  whilst  the 
pits  of  I2  and  I3  are  almost  twinned  near  the  sagittal  line.  Ix  is  slightly  raised, 
especially  anterolaterally,  and  the  pit-count  is  as  follows  :  E:  29  or  29  estd  ;  Ir  22  ; 
I?22:  I322;  I4  20  (from  R4)  ;  1 5  estd  17  or  18  from  about  R6  ;  I6  estd  16  from  Rio. 
It  is  not  evident  whether  I7  is  represented.  Fragmentary  evidence  of  Marrolithoides 
was  also  obtained  from  locality  B .  18. 

Genus  CRYPTOLITHUS  Green,  1832 

Cryptolithus?  inferus  sp.  nov. 

(PL  2,  figs.  2,  4,  6-8,  12) 

Diagnosis.  Cryptolithus?  with  cephalon  about  twice  as  broad  as  long,  its  outline 
subangular  anterolaterally  and,  occasionally,  frontally.  Cephalic  fringe  narrow 
(sag.)  anteriorly  with  only  three  (E1;  ^-2),  rarely  two,  concentric  rows  of  pits 
continuous  in  front  of  glabella.  I3  generally  developed  from  about  R4.  Several 
intercalated  pits  in  Ex,  especially  frontally  and  anterolaterally.  Pit  count  relatively 
low,  as  follows  :  E1  +  ex  27-30  ;  Ix  -|-  ix  22-26  ;  I2  -f  i2  21-22  ;  I3  -f  i3  19  ;  I4  +  i4 
16-17  1  I5  comprises  about  a  dozen  small  pits  in  region  of  Rio  to  R18  ;  I6  not 
developed.     Reticulation  of  test  weak  or  absent  in  large  cephala. 

Holotype.     It. 734  (PI.  2,  fig.  8). 

Paratype.     It. 735  (PI.  2,  fig.  6). 

Other  material.  It. 689  (PL  2,  fig.  12)  ;  It. 697  (PL  2,  fig.  7)  ;  It. 703  (PL  2, 
fig-  4)  ;   It. 704  (PL  2,  fig.  2). 

Localities  and  horizons.  The  holotype  and  paratype  are  from  locality  B .  1, 
south-west  of  Bedinan,  where  a  sample  of  thirteen  specimens  was  obtained  from  the 
lowest  fossiliferous  portion  of  the  Bedinan  Formation  seen  there.  Another  sample, 
also  comprising  thirteen  specimens,  from  B.2,  a  little  higher  in  the  succession  shows 
a  pit  count  for  Ex  +  ex  which  is  consistently  higher  than  that  for  the  B .  1  material 
(31-34  pits  compared  with  27-30)  and  is  here  termed  C.  cf.  inferus  (see  below). 
The  specimens  from  the  two  localities  are  otherwise  indistinguishable.  A  single 
cranidium  of  C?  cf.  inferus  (PL  2,  fig.  12)  was  found  at  B.3,  associated  with 
Marrolithoides  laticirrus. 

Description.  The  cephalon  is  about  twice  as  broad  as  long,  sometimes  slightly 
broader,  especially  in  smaller  specimens,  and  its  outline  is  generally  well  rounded 
except  frontally,  where  it  may  be  almost  subangular.  The  glabella  is  high,  narrow 
and  relatively  long,  occupying  up  to  five-sixths  or  more  of  the  cephalic  length 
(excluding  occipital  spine).  It  expands  forwards  only  slightly  to  the  rounded 
frontal  glabellar  lobe,  which  extends  a  little  in  front  of  the  cheek-lobes,  and  the  sides 
are  bounded  by  almost  straight  axial  furrows.  The  latter  contain  a  pair  of  elongated 
alar  lobes  in  immature  specimens,  but  these  structures  diminished  during  ontogeny 
and  are  absent  from  the  presumed  adult  stages.  Small  specimens  also  exhibit 
reticulation  of  the  test  of  the  cheek-lobes  and  centre  of  the  glabella,  but  in  the  largest 
examples  from  the  type  locality  this  is  weak  or  absent.     On  the  other  hand  several 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  103 

specimens  from  B.2  show  stronger  reticulation,  but  this  may  be  no  more  than  an 
accident  of  preservation 

The  cephalic  fringe  is  only  moderately  broad,  especially  frontally  but  also,  to  a 
lesser  degree,  anterolaterally  and  laterally.  Frontally  it  is  constricted  slightly  by 
the  extension  forwards  of  the  glabella  and  contains  three  concentric  rows  of  pits 
(E1(  I1-2).  Ex  extends  to  the  genal  angles  and  includes  several  intercalated  pits, 
especially  frontally  but  also  anterolaterally  and  laterally.  The  pits  of  Et  decrease 
slightly  in  diameter  towards  the  genal  angles,  and  frontally  they  are  of  approxi- 
mately similar  size  to  those  of  I2.  The  pits  of  Ij  are  larger  than  those  of  other 
rows  and  they,  too,  diminish  a  little  anterolaterally.  I3  is  developed  from  R3  or 
R4,  I4  from  about  R9,  and  I5  from  about  Rio  to  R12.  The  following  pit  count 
is  based  on  thirteen  specimens  from  B .  1  :  Ex  +  ex  27-30  ;  Ix  -j-  ix  22-26  ; 
I2  +  i2  22  ;  I3  +  i3  19  ;  I4  +  i4  16.  In  the  holotype  I5  consists  of  12  small  pits 
which  extend  from  Rio  to  R17  or  R18  ;  the  row  then  terminates  until  just  in  front 
of  the  posterior  border  furrow,  where  three  further  pits  form  an  apparent  continua- 
tion. The  hindmost  of  these  three  is  slightly  the  largest  and  probably  represents  the 
position  of  an  apodeme,  sited  at  the  fulcrum  and  functional  in  the  articulation  of 
posterior  border  and  first  thoracic  segment.  The  material  available  is  insufficient 
to  show  whether  this  break  in  the  line  of  1 5  is  a  reliable  specific  character,  but  a 
broadly  similar  break  was  found  in  at  least  two  other  specimens  at  the  type  locality. 
A  sample  of  thirteen  specimens  from  B.2,  some  10  m.  higher  in  the  succession  (see 
Text-fig.  3),  gave  the  following  pit  count :  Ex  +  ex  31-34  ;  Ii  +  ii  22-23  '•  I2  +  h 
21-22  ;  I3  +  i3  19  (from  R2)  ;  I4  +  h  17  (from  R4).  Again,  I5  is  discontinuous, 
with  twelve  to  thirteen  pits  anterolaterally  and  a  further  two  or  three  near  the 
posterior  border  furrow.  The  slightly  higher  number  of  pits  in  Ex  is  not  considered 
to  justify  separation  of  these  specimens,  which  are  listed  and  figured  as  Cryptolithus? 
cf.  inferus  (PL  2,  figs.  2,  4,  7).  The  marginal  cephalic  suture  is  of  normal  trinucleid 
type  and  at  the  genal  angles  the  lower  fringe  lamella  is  produced  posterolaterally  to 
form  a  pair  of  librigenal  spines  which  are  long  and  slender,  at  least  twice  the  median 
length  of  the  cephalon,  prismatic  in  cross-section,  and  have  a  longitudinal  ridge 
continuous  with  the  EJI-l  girder.  In  addition,  two  pseudogirders  (IJI2  and  I2/I3) 
are  fairly  well  developed,  particularly  frontally. 

The  thorax  consists  of  six  segments,  the  first  of  them  slightly  macropleural.  The 
axis  occupies  about  one-quarter  of  the  total  breadth,  stands  a  little  higher  than  the 
side-lobes,  and  is  bounded  by  shallow,  broad  axial  furrows.  Each  segment  has  a 
small,  articulating  half-ring  separated  from  the  axial  ring  by  an  articulating  furrow 
which,  on  the  internal  mould,  appears  deep  and  broad  (sag.)  with  a  pair  of  apodemes 
sited  abaxially.  The  pleurae  are  horizontal,  parallel-sided,  their  tips  obliquely 
truncated  to  form  posterolateral  points.  Each  pleura  carries  a  broad,  shallow, 
pleural  furrow  which  runs  almost  straight  from  the  anterior  margin,  at  the  axial 
furrow,  and  just  fails  to  attain  the  pleural  tip. 

The  pygidium  is  closely  similar  to  that  found  in  other  species  of  the  genus,  that  is 
is  to  say  its  outline  is  an  Isosceles  triangle,  three  and  a  half  times  as  broad  as  long, 
with  a  transversely  straight  frontal  margin  and  broadly  divergent,  faintly  sinuous, 
lateral  margins.     The  small  axis  occupies  one-sixth  of  the  frontal  breadth,  tapers 


104  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

back  to  the  tip,  and  is  separated  from  the  flat  side-lobes  by  gently  curved,  shallow 
axial  furrows.  There  are  about  ten  axial  rings,  those  after  the  sixth  being  less  well 
defined.  The  side-lobes  carry  four  and  a  half  pairs  of  more  or  less  well-defined 
pleural  ribs,  occasionally  with  traces  of  a  fifth  pair. 

A  discussion  of  Cryptolithusl  infer  us  and  other  species  of  the  genus  follows  the 
description  of  C?  bedinanensis  (see  below). 

Cryptolithus  ?  bedinanensis  sp.  nov. 
(PL  3,  figs.  1-7,  9  ;  PI.  4,  figs.  2-9) 
Diagnosis.  Cryptolithus  with  cephalon  (excluding  librigenal  spines)  about 
twice  as  broad  as  long,  its  outline  rounded  frontally,  but  with  sides  almost  straight 
and  slightly  convergent  forwards.  Cephalic  fringe  moderately  broad,  typically  with 
3  concentric  rows  of  pits  (E1;  I]_2)  in  front  of  glabella,  and  traces  of  ridge  between  Ex 
and  I-t  anterolaterally.  I3  well  developed,  and  proportion  of  specimens  with  4  rows 
of  pits  in  front  of  glabella  increases  higher  in  stratigraphical  succession.  I^g  pits 
larger  than  those  of  adjacent  rows.  Ex  extends  to  genal  angles  and  contains  several 
intercalated  pits.  Pit  count  as  follows  :  Ex  typically  32-36  but  full  range  26-38  ; 
Ix  21-27,  mostly  23-27  ;  I2  21-26,  mostly  23-25  ;  I3  20-25,  mostly  22-25  »'  I4  I6-23, 
mostly  18-21  ;  I5  10-18,  mostly  14-18  ;  I6  6-13,  mostly  7-10  ;  I7  rarely  seen,  with 
only  very  few  pits.  Thorax  and  pygidium  of  general  cryptolithinid  type  ;  latter 
has  about  ten  axial  rings  and  five  and  a  half  pairs  of  pleural  ribs. 

HOLOTYPE.       It.I2I0  (PI.  3,  figs.  2,  3). 

Paratypes.     It.1211  (PL  3,  fig.  6)  ;   It.1231  (PL  3,  fig.  9). 

Localities  and  horizons.  The  lowest  stratigraphical  occurrence  of  the  forma 
typica  is  at  B.6,  south-west  of  Bedinan,  where  the  largest  sample  (29  specimens)  was 
obtained.  This  is  also  the  type  locality,  and  the  greater  part  of  the  sample  showed  3 
pit  rows  in  front  of  the  glabella,  though  a  few  had  4  rows  (see  PL  3,  fig.  1).  The 
species  was  found  subsequently  throughout  the  remainder  of  the  mudstone  sequence 
of  the  Bedinan  Formation  and  the  proportion  of  specimens  with  4  frontal  rows  of 
pits  increased  until  the  ratio  of  the  two  types  at  B.6  was  almost  reversed  at  B.16 
(for  data,  see  below). 

Description.  The  entire  dorsal  exoskeleton  of  a  slightly  compressed  individual 
is  a  little  broader  than  long,  approximately  in  the  ratio  7:6.  Just  over  half  the 
median  length  is  occupied  by  the  cranidium,  which  is  slightly  more  than  twice  as 
broad  as  long,  the  maximum  breadth  being  measured  across  the  genal  angles.  The 
outline  is  generally  well  rounded  frontally  and  anterolaterally,  but  often  straighter 
towards  the  genal  angles.  The  glabella  expands  forwards  gently  in  both  height  and 
breadth  for  about  five-sixths  of  its  length  and  then  contracts  to  form  a  well-rounded 
frontal  lobe.  It  is  set  higher  than  the  convex  cheek-lobes  and  is  separated  from  them 
by  broad,  nearly  straight,  axial  furrows.  The  latter  become  slightly  broader 
posteriorly  in  the  adult  trilobite,  and  even  more  so  in  immature  individuals  so  as  to 
accomodate  a  pair  of  low,  elongated  alar  lobes.  As  in  other  cryptolithinids  the  alar 
lobes  became  progressively  smaller  during  ontogeny  and  eventually  disappeared. 
The  sides  of  the  glabella  carry  two  pairs  of  shallow  impressions  which  represent 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  105 

glabellar  furrows  delimiting  two  small  pairs  of  almost  indiscernible  glabellar  lobes 
(PL  3,  fig.  9).  A  shallow  occipital  furrow  deepens  abaxially  to  where,  as  seen  on 
the  internal  mould,  a  pair  of  apodemes  is  sited  behind  the  outer  margins  of  the  axial 
furrows.  The  occipital  ring  is  short  {sag.),  slopes  backwards  gently,  and  is  produced 
to  form  a  slim  occipital  spine.  The  quadrant-shaped  cheek-lobes  extend  forwards  as 
far  as  the  line  of  maximum  breadth  of  the  glabella.  Both  they  and  the  axial  portion 
of  the  glabella,  which  is  surmounted  by  a  small,  median  tubercle,  frequently  carry  a 
fine,  mesh-like  ornamentation  of  raised  ridges,  though  this  is  not  always  preserved 
and  may  not  always  have  persisted  in  large  individuals.  In  one  or  two  immature 
examples  a  pair  of  fine  nervures  is  visible  on  the  cheek-lobes.  The  posterior  border 
furrow  is  transversely  straight,  of  only  moderate  depth,  and  becomes  broader 
(exsag.)  abaxially  ;  its  posterior  margin  is  more  steeply  inclined  than  the  anterior. 
The  posterior  border  is  narrow  (exsag.),  ridge-like,  transversely  straight  for  less  than 
half  the  distance  from  the  axial  furrows  to  the  lateral  margins  ;  it  then  meets  a  pair 
of  fulcra  and  flexes  down  and  slightly  backwards  around  relatively  small  posterior 
prolongations  of  the  cephalic  fringe.  In  most  trinucleids  the  cephalic  fulcra  are 
denoted  by  a  pair  of  pits,  often  large,  in  the  posterior  border  furrow.  In  the  case  of 
Cryptolithus}  bedinanensis  it  is  not  clear  whether  they  correspond  to  the  hindmost 
pits  of  the  innermost  concentric  row  of  the  fringe. 

The  cephalic  fringe  is  of  moderate  breadth  frontally,  where  3  or  4  concentric  rows 
of  pits  (Ex  and,  respectively,  Ix_2  or  I^g)  are  developed  in  front  of  the  glabella. 
The  fringe  becomes  broader  anterolaterally  and  laterally,  where  additional  I  rows 
soon  appear.  Many  specimens  show  a  low,  thin  ridge  on  the  dorsal  lamella,  running 
between  ~EX  and  Ix,  and  diminishing  frontally  and  posterolaterally.  It  is  not  clear 
whether  this  was  a  primary  structure,  but  such  a  ridge  could  well  have  been  formed 
by  dorsal  compression  of  a  fringe  in  which  Ix  was  originally  set  slightly  higher  than 
the  adjacent  rows,  particularly  anterolaterally  (see  PL  4,  fig.  6).  In  the  case  of 
cephala  with  3  concentric  pit  rows  at  the  sagittal  line,  the  pitted  area  is  invaded  to  a 
greater  degree  by  the  front  of  the  glabella,  and  it  was  thought  at  first  that  there  were 
grounds  for  separating  such  forms  from  others  possessing  4  complete  frontal  rows 
and  little  extension  forwards  of  the  glabella.  All  now  appear  to  fall  within  the 
limits  of  variation  for  the  species  but  the  proportion  of  specimens  with  4  rows  (even 
in  small  cephala)  increases  as  one  ascends  the  succession,  as  shown  below. 


No. 

with 

3 

No. 

with  4 

ocality 

Sample 

frontal  rows 

frontal  rows 

B.6 

29 

18 

2 

or  3 

B.8 

6 

1 

2 

B.g 

1 

- 

1 

B.io 

4 

2 

1 

B.n 

2 

- 

1 

B.12 

6 

4 

2 

B.13 

13 

3 

6 

B.14 

3 

1 

1 

B.15 

11 

2 

5 

B.16 

16 

2 

12 

B.20 

1 

- 

- 

B.22 

4 

- 

1 

io6 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.    TURKEY 


When  three  frontal  rows  of  pits  are  present  the  succeeding  row,  13,  is  always  well 
developed  and  may  extend  as  far  forwards  as  R3  or  R4.  Of  the  remaining  rows,  I4 
extends  from  R5  or  R6  (occasionally  R8  or  R9),  I5  from  R7  to  Rg  (less  commonly 
R9  to  R15),  whilst  I6,  although  not  always  present  in  immature  cephala,  is  usually 
found  as  far  forwards  as  R12  to  R15.  I7  is  represented  in  only  two  specimens,  from 
B.13  and  B.16,  by  five  small  pits  along  the  anterolateral  boundary  of  each  cheek- 
lobe.  There  is  only  one  E  row,  extending  to  the  genal  angles  and  composed  of 
numerous  pits  which  greatly  outnumber  those  of  the  I  rows  and  are  correspondingly 
more  difficult  to  fit  into  a  radial  pattern.  The  I  rows  exhibit  a  strong  radial  arrange- 
ment which  persists  to  within  a  few  (generally  3  or  4)  pits  of  the  posterior  border ; 
usually  there  are  17  to  21  radii,  though  occasionally  up  to  23  in  individuals  with  a 
particularly  high  overall  pit-count.  The  pits  of  Ix  and  I?  are  of  about  equal  size, 
somewhat  larger  than  those  of  Ex  and  I2  which  are  also  about  equisized  ;  the  pits  of 
the  remaining  rows  become  smaller  from  I4  to  I6.  There  are  occasional  deviations 
from  this  general  rule  and  the  pits  of  Ex  are  sometimes  slightly  smaller  than  those  of 
I3.  The  following  table  shows  the  variation  in  the  number  of  pits  present  in 
individual  concentric  rows  on  the  cephalic  fringe  of  Cryptolithus?  bedinanensis.  A 
blank  indicates  that  the  material  available  was  insufficiently  well-preserved  to 
obtain  a  reliable  figure.  At  B.6  and  B.13  the  wide  range  of  variation  is  accounted 
for  by  small  cephala  with  a  lower  pit-count. 


No.  in 
Locality     sample    K1  +  e2    Ix  +  ix     I2  +  i2     I3 


I*  +  14     Is  +  is     I6  +  h 


I7 


B.6 


29 


26-36       23-27       21-26       21-25       16-19       10-14       c-  6-8 


esp. 

esp. 

esp. 

34-36 

23-25 

23-25 

B.8 

6 

27-31 

21 

22 

22-23 

18-19 



c.  10 

— 

B.9 

1 

29 

c.  22 

— 

— 

— 



— 

— 

B.io 

4 

28-33 

22-24 

21-23 

22 

18 

17 

c.  10 

— 

B.n 

2 

33 

23 

c.  24 

c.  21 

— 



— 

— 

B.12 

6 

29-35 

25-26 

24-26 

22-25 

13-18 

c.  17 

c.  7 

— 

B.13 

13 

29-38 

esp. 

33-38 

21-27 

esp. 

23-27 

21-26 

esp. 

24-26 

22-25 

18-23 

17-18 

6-7 

5  pits 

in  one 

specimen 

B.14 

3 

32-34 

23-24 

23-25 

20 

— 



— 

— 

B.15 

11 

32-35 

23-25 

23-25 

22-25 

20-21 

16-18 

— 

— 

B.16 

16 

31-38 

22-26 

22-26 

22-25 

20-2I 

c.  16 

5-6 

5  pits 

in  one 

specimen 

B.20 

1 

— 

c.  24 

— 

c.  20 



— 

— 

— 

B.22 

4 

3i 

c.  23 

c.  23 

23 

20-2I 

c.  16 

12-13 

— 

On  the  ventral  lamella  of  the  cephalic  fringe  each  genal  angle  is  produced  postero- 
laterally  to  form  a  long,  slim,  gently-curved  librigenal  spine.  Each  spine  is  prismatic 
in  cross-section  with  longitudinal  ridges,  the  lower  of  which  extends  a  little  way  into 
the  pitted  area  of  the  fringe  and  then  bifurcates.  Of  the  branches  so  formed,  one  is 
developed  as  a  thin,  sharp  ridge,  the  true  girder,  between  Ex  and   I1  ;    the  other 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  107 

forms  a  slightly  broader  but  lower  ridge,  the  IJI2  pseudogirder.  A  further  pseudo- 
girder  is  developed  between  I2  and  I3  both  frontally  and  anterolaterally,  where  it 
attains  approximately  the  same  dimensions  as  the  IJl^  pseudogirder.  These 
structures  are  shown  clearly  in  PL  3,  fig.  7. 

A  few  immature  specimens  of  Cryptolithus?  were  collected  from  the  Bedinan 
Formation  and  one  of  the  smallest  and  best  preserved  is  shown  in  PL  3,  fig.  8.  It 
exhibits  the  features  characteristic  of  a  trinucleid  Meraspis,  including  genal  angles 
set  in  front  of  the  line  of  the  posterior  border  furrow,  and  well-developed  alar  lobes. 
The  fringe  is  narrow,  consisting  almost  solely  of  two  concentric  pit  rows,  but  the 
count  of  23  pits  is  high. 

The  thorax  consists  of  six  segments,  the  first  one  macropleural,  and  is  indistin- 
guishable from  other  members  of  the  genus.  Material  from  the  type  locality  shows 
clearly  that  the  pleural  tips  are  blunt,  almost  vertical,  and  form  a  virtual  continuation 
of  the  form  of  the  anterolateral  margins  of  the  pygidium  (see  PL  3,  fig.  2). 

The  pygidium  is  also  of  the  form  characteristic  for  the  genus,  with  a  low,  narrow, 
marginal  rim.  The  margin  is  moderately  declined  over  the  median  third  but 
becomes  vertical  anterolaterally  (see  above).  In  apparently  adult  examples  the 
axis  has  the  first  four  rings  well  defined,  with  only  traces  of  a  further  six  rings. 
Small  specimens  generally  have  better-defined  axial  rings  which  appear  confluent 
with  the  pleural  ribs  (for  example  PL  3,  fig.  6,  probably  Meraspis,  Degree  5).  The 
side-lobes  carry  about  five  and  a  half  pairs  of  ribs,  separated  by  broad  (exsag.), 
shallow,  pleural  furrows  which  broaden  towards  the  margin. 

A  cranidium  of  unusual  type  (PL  2,  fig.  10)  was  found  at  locality  B  .6  in  association 
with  Cryptolithus?  bedinanensis.  It  differs  from  the  latter  in  having  a  particularly 
short  glabella  and  lateral  margins  which,  judging  from  the  incomplete  material, 
must  have  been  slightly  divergent  forwards.  The  specimen  is  apparently  malformed 
and  therefore  difficult  to  compare  with  other  species.  The  concentric  arrangement 
of  pits  is  relatively  irregular  frontally,  where  I2  is  of  normal  type,  Ix  is  deflected 
slightly  inwards,  and  Ex  is  displaced  markedly  inwards  from  the  margin.  Conse- 
quently several  pits,  estimated  as  6  or  7  on  the  complete  cephalon,  are  arranged  in 
interradial  positions  frontally,  almost  as  an  incipient  E2.  The  specimen  is  listed 
merely  as  Cryptolithus?  sp. 

Another  example  of  a  pathological  cranidium  was  found  at  B .  13  and  is  figured  as 
Cryptolithus?  cf.  bedinanensis  (PL  4,  fig.  1).  The  left  side  generally  resembles  that 
species  but  on  the  right  side  E2  does  not  extend  as  far  forwards  as  the  axial  furrow, 
whilst  Ix  and  I2  become  irregular  and  flex  forwards,  the  former  row  reaching  the 
cranidial  margin  in  front  of  the  right  axial  furrow. 

Discussion.  All  the  species  of  Cryptolithus  described  by  Whittard  (1958  :  72 
et  seq.)  from  the  Shelve  Inlier  were  obtained  from  strata  of  the  Llanvirn  and 
Llandeilo  Series  and  so  are  considerably  older  than  the  Turkish  specimens.  The 
Anglo-Welsh  forms  possess  cheek  lobes  which  are  relatively  large  and  extend 
forwards  level  with  the  front  of  the  glabella,  whereas  in  the  Turkish  material  the 
glabella  invariably  extends  beyond  the  cheek-lobes  and  often  invades  the  pitted 
fringe  area,  to  be  accomodated  by  a  flexing  forwards  of  the  inner  concentric  rows  of 
pits.     Another  conspicuous  point  of  difference  is  that  the  Anglo-Welsh  specimens 


io8  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

exhibit  an  unusually  large  range  of  pit  size,  those  of  the  inner  and  posterolateral 
parts  of  the  fringe  being  often  exceedingly  small ;  The  Turkish  specimens  have  the 
pits  of  Ix  and  I2  of  approximately  similar  size,  whilst  the  pits  of  all  the  other  rows, 
although  smaller,  are  never  minute.  An  exception  to  these  generalizations  is 
offered  by  the  cephalon  of  "  Cryptolithus  sp.  "  figured  by  Whittard  (1956,  pi.  9, 
fig.  16)  from  the  Lower  Llanvirn  of  the  Llandrindod  area  of  Wales.  In  this  form  the 
pits  are  both  fewer  and  of  simpler  disposition,  features  more  reminiscent  of  the 
North  American  species  of  the  genus,  but  only  one  example  is  yet  known.  Some 
of  the  specimens  figured  by  Whittard  have  relatively  large  pygidia  and  in  one 
individual  (Whittard  1956,  pi.  10,  fig.  5)  referred  to  Cryptolithus  cf.  inopinatus  the 
pygidium  seems  atypical  of  that  genus  but  strongly  suggestive  of  Lloydolithus.  It 
is,  of  course,  possible  that  groups  of  trinucleids  with  rounded  cephalic  outline  and 
one  row  of  pits  external  to  the  girder  arose  independently  from  different  genera,  and 
may  also  have  been  differentiated  geographically,  but  the  evidence  is  not  conclusive. 

A  feature  of  all  the  specimens  of  Cryptolithus?  from  the  Bedinan  Formation  is  the 
development  of  Ex  as  a  row  of  numerous  small  pits  extending  to  the  genal  angle. 
This  is  reminiscent  of  the  dorsal  aspect  of  the  cephalic  fringe  in  Onnia  Bancroft 
from  the  higher  Caradoc  Series  of  South  Shropshire  (for  discussion  see  Dean  i960  : 
127),  but  in  the  latter  case  the  pits  of  E1  and  I±  are  larger  than,  and  raised  above, 
those  of  adjacent  rows,  whilst  radial  sulci  are  well  developed  from  I2  inwards.  The 
ventral  fringe  surface  of  Onnia  is  particularly  distinctive,  having  a  strong  girder  with 
two  concentric  rows  of  pits  (E^g)  external  to  it.  The  girder  itself  is  continuous  with 
a  ridge  along  the  librigenal  spine  and  there  is  a  lesser  development  of  pseudogirders 
between  concentric  rows  internal  to  the  girder,  the  IJI2  pseudogirder  at  least  arising 
from  a  subsidiary  branch  of  the  main  structure  of  ridge  and  girder.  A  somewhat 
similar  condition  is  found  in  the  Turkish  specimens  but  with  only  a  single  E  row 
developed.  In  addition  some  examples  have  the  girder  more  weakly  developed  so 
that  it  appears  scarcely  more  prominent  than  the  adjacent  pseudogirders. 

Cryptolithus?  infer  us  shows  only  relatively  small  differences  from  the  majority  of 
the  stratigraphically  earlier  members  (including  the  types)  of  C?  bedinanensis , 
that  is  to  say  specimens  which  also  have  three  concentric  rows  of  pits  in  front  of  the 
glabella.  C?  inferus  typically  has  a  much  lower  Ex  +  et  count  ;  there  is  one  fewer  I 
row  (this  is  particularly  noticeable  anterolaterally,  where  the  fringe  is  narrower)  ; 
and  the  anterior  portion  of  Ej  consists  of  slightly  larger  pits  with  fewer  intercalated 
pits. 

One  can  postulate  an  evolutionary  sequence  within  the  members  of  Cryptolithus? 
in  the  Bedinan  Formation.  The  earliest  representatives,  typical  C?  inferus,  soon 
acquire  a  slightly  higher  number  of  pits  in  Ex  (the  latter  specimens  are  denoted  as 
C?  cf.  inferus).  Higher  in  the  succession  appears  C?  bedinanensis  which  could  have 
developed  from  C?  inferus  by  the  addition  of  extra  pits  in  almost  all  rows,  and  one 
concentric  row  internal  to  the  girder.  Further  increase  in  the  pits  of  I3  led  to  the 
formation  of  four  frontal  rows  in  the  majority  of  later  members  of  the  species,  though 
the  average  count  for  individual  rows  does  not  apparently  show  any  significant 
increase. 

Judging  from  the  marked  Bohemian  affinities  of  the  remainder  of  the  fauna,  one 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  109 

might  expect  to  compare  the  Bedinan  trinucleids  with  those  of  Bohemia,  but  the 
lack  of  relevant  literature  makes  this  difficult.  Whittington  (1940)  considered  the 
Czech  species  Trinncleus  ornatus  Sternberg,  T.  goldfussi  Barrande  and  T.  ultimus 
Barrande  to  be  identical,  and  redescribed  them  as  Onnia  ornatus  (Sternberg).  The 
type  material  of  all  these  species  comes  from  different  stratigraphical  horizons  and 
Marek  (1952  :  23)  has  pointed  out  that  Trinucleus  ultimus,  at  least,  merits  specific 
separation.  Of  the  material  illustrated  by  Whittington,  his  pi.  3,  fig.  6  bears  some 
resemblance  to  Cryptolithus?  bedinanensis  with  three  frontal  pit  rows  (this  paper, 
PI.  3,  fig.  3),  but  the  pits  of  the  outermost  row  seem  fewer  and  rather  more  regularly 
arranged  in  the  Czech  specimen.  The  ventral  fringe  surface  illustrated  by  Whitting- 
ton (1940,  pi.  3,  fig.  5)  shows  a  girder  and  pseudogirders  which  are  perhaps  more  like 
those  of  the  Turkish  specimens  than  in  Onnia  (s.s.).  The  original  of  Whittington's 
pi.  3,  fig.  6  was  said  to  have  been  labelled  as  Trinucleus  goldfussi  by  Barrande,  and 
the  preservation  in  a  fine-grained,  buff  sandstone  (Whittington  1940  :  243)  suggests  a 
possible  origin  in  the  Letna  Beds  of  the  Caradoc  Series.  The  specimen,  a  cranidium, 
has  four  frontal,  concentric  rows  of  pits,  several  rows  anterolaterally,  and  the 
lateral  margins  are  slightly  divergent  forwards.  The  outline  is  reminiscent  of  some 
examples  of  Marrolithoides  laticirrus,  but  the  latter  has  a  much  larger  number  of 
pits  in  the  outermost  concentric  row.  Further  discussion  of  these  problems  must 
await  a  modern  revision  of  the  Bohemian  trinucleids. 


Family  DIONIDIDAE  Gurich,  1907 

Genus  DIONIDE  Barrande,  1847 

Dionide  formosa  (Barrande)  anatolica  subsp.  nov. 

(PL  5,  figs.  1-4,  6,  7,  12) 

Diagnosis.  Subspecies  of  Dionide  formosa  distinguished  by  following  features  : 
cephalic  fringe  longer  (sag.)  in  front  of  glabella,  with  less  well-developed  marginal 
row  of  fringe-pits  ;  genal  angles  turned  backwards  less  strongly  ;  pygidium 
proportionately  shorter  with  slightly  narrower  axis,  eleven  to  thirteen  axial  rings, 
and  ten  to  twelve  pairs  pleural  ribs. 

Holotype.     It.1061  (PI.  5,  figs.  1,  2). 

Paratypes.  It.  1057  (PL  5,  fig.  12)  ;  It.  1062  (PL  5,  fig.  6)  ;  It.  1063  (PL  5, 
fig.  7)  ;   It.1178  (PL  5,  fig-  4)  ;   It.1201  (PL  5,  fig.  3). 

Horizons  and  localities.  All  the  type  material  is  from  the  section  near  Ziyaret, 
east  of  Sosink,  the  holotype  being  collected  from  locality  A .  3  and  the  paratypes  from 
A. 3,  A. 5  and  A. 6.  All  these  localities  are  believed  to  be  in  the  lower  part  of  the 
Bedinan  Formation. 

Description.  The  cephalon,  excluding  librigenal  spines,  is  transversely  almost 
semielliptical  in  plan  ;  the  median  length  varies  from  just  over  one-third  to  almost  a 
half  of  the  maximum  breadth,  according  to  the  degree  of  crushing  of  the  specimens. 
The  tumid  glabella  is  roughly  subquadrate  in  outline,  the  frontal  lobe  rounded, 


no  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

convex  forwards  and  bounded  by  a  narrow  (sag.),  deep  preglabellar  furrow.  The 
posterolateral  portions  of  the  glabella  are  formed  by  a  pair  of  large  lobes,  almost 
reniform  in  plan,  which  project  slightly  beyond  the  lateral  margins  of  the  glabella 
and  are  set  below  the  level  of  the  median  body.  They  are  defined  by  a  pair  of  broad, 
shallow  furrows  which  deepen  posteriorly.  One  specimen  (PL  5,  fig.  4)  shows  the 
hindmost  part  of  the  glabella  forming  a  depressed,  neck-like  structure,  its  anterior 
boundary  immediately  behind  the  posterior  ends  of  the  lateral  lobes.  As  all  the 
material  is  crushed,  it  is  not  clear  whether  this  structure  is  primary.  A  narrow 
occipital  furrow  separates  the  glabella  from  a  small,  occipital  ring  which  becomes 
shorter  (exsag.)  towards  the  axial  furrows.  Deep  axial  furrows  separate  the  glabella 
from  cheek-lobes  which  are  almost  quadrant-shaped,  slightly  broader  than  long. 
The  cranidium  is  circumscribed  by  a  narrow,  low,  marginal  rim  which  is  ridge-like 
frontally  but  becomes  slightly  broader  and  less  inclined  posterolaterally.  Internal 
to  the  rim  is  a  pitted  fringe  of  the  type  characteristic  for  the  genus,  and  this  attains  a 
breadth  (sag.)  frontally  of  between  one-quarter  and  one-third  of  the  length  of  the 
glabella  ;  this  refers,  of  course,  to  compressed  material  and  takes  no  account  of  any 
original  inclination  of  the  fringe.  Laterally  the  line  of  demarcation  between  the 
fringe  and  cheek-lobes  is  difficult  to  distinguish,  owing  to  crushing  and  the  fact  that 
both  carry  similar  ornamentation.  A  comparable  state  of  affairs  was  noted  by 
Whittard  (1958  :  99)  when  redescribing  Dionide  jubata  Raymond.  The  cheek-lobes 
are  separated  by  a  broad  (exsag.),  posterior  border  furrow  from  a  posterior 
border  which  is  at  first  transversely  straight  and  almost  uniformly  broad  but  then 
becomes  narrower,  turns  back  slightly,  and  finally  coalesces  with  the  hindmost, 
less  inclined  parts  of  the  marginal  rim  described  earlier.  The  surface  of  each  cheek- 
lobe  is  traversed  by  a  conspicuous,  branched  nervure  which  runs  posterolaterally 
from  the  axial  furrow,  opposite  the  centre  of  the  glabella,  towards  the  genal  angle. 
The  nervure  comprises  thickened,  irregular  ridges  whose  development  is  somewhat 
variable.  In  two  cases  (PL  5,  figs.  4,  6)  each  nervure  includes  two  branches  which 
coalesce  near  the  genal  angle  and  then  die  out  quickly.  In  another  (PL  5,  fig.  3) 
only  a  single  ridge  is  visible,  whilst  the  holotype  (PL  5,  figs.  1,  2)  shows  a  third, 
smaller  branch  in  front  of  the  main  pair,  the  anterior  of  which  is  slightly  the  thicker. 
The  glabella  is  surmounted  by  a  conspicuous  median  spine,  rounded  in  cross-section 
and  directed  backwards  and  slightly  upwards.  In  the  somewhat  crushed  holotype 
the  length  of  the  spine  approximates  to  that  of  the  cranidium,  but  it  was  probably 
even  longer  originally.  A  similar  structure  in  Dionide  formosa  was  illustrated  by 
Hawle  &  Corda  (1847,  pi.  3,  fig.  16)  but  is  not  usually  preserved.  More  recently, 
however,  Curtis  (1961  :  14)  has  noted  a  Portuguese  example  with  the  spine  at  least 
5  mm.  long.  One  specimen  from  the  Bedinan  Formation  shows  the  left  librigena 
(PL  5,  fig.  3)  which,  although  damaged,  is  seen  to  end  in  a  long,  curved  librigenal 
spine,  the  original  length  of  which  is  estimated  to  have  been  at  least  twice  that  of  the 
cephalon. 

A  complete  thorax  has  not  been  found  but  one  incomplete  specimen  (PL  5,  fig.  3) 
shows  four  segments,  the  first  of  them  markedly  macropleural.  Another  (PL  5, 
fig.  7),  with  pygidium  attached,  has  five  equisized  segments  and  is  also  incomplete 
as  it  lacks  a  macropleural  segment.     It  is  reasonable  to  assume  that  six  segments 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  in 

would  normally  be  present,  the  customary  complement  for  the  genus.  The  axis 
occupies  just  over  one-fifth  of  the  thoracic  breadth,  stands  only  a  little  higher  than  the 
side-lobes,  and  is  bounded  by  narrow,  almost  straight,  axial  furrows  which  converge 
gently  backwards.  Each  axial  ring  is  almost  rectangular  in  plan  but  ends  antero- 
laterally  in  a  well-defined  pair  of  subtriangular  axial  lobes  which  are  more 
conspicuous  on  the  internal  mould.  The  pleurae  are  transversely  straight,  parallel- 
sided,  and  terminate  in  blunt  points  directed  posterolaterally.  The  breadth 
(exsag.)  of  the  pleurae  of  the  first,  macropleural  segment  is  about  one-third  that  of  the 
remaining  segments.  A  pleural  furrow  runs  from  the  anterior  edge  of  each  pleura 
immediately  outside  the  axial  furrow.  At  first  it  is  deep  and  narrow,  and  runs 
almost  to  the  centre  of  the  pleura  in  a  broad  curve,  concave  forwards  ;  it  then 
becomes  broader  (exsag.)  and  shallower,  and  runs  backwards  very  slightly,  sub- 
parallel  to  the  anterior  margin,  almost  to  the  pleural  tip,  where  it  turns  back  more 
strongly  and  dies  out.  In  the  case  of  the  macropleural  first  segment  the  pleural 
furrows  are  directed  backwards  more  strongly,  and  the  inner  part  of  the  anterior 
band  (PI.  5,  fig.  3)  is  inflated,  as  in  Dionide  formosa  formosa  (see  Whittington  1952, 
Text-fig.  1). 

The  best-preserved  pygidium  (PL  5,  fig.  7)  is  slightly  more  than  three  times  as 
broad  as  long,  sub-semielliptical  in  outline  with  the  anterior  margin  almost  trans- 
versely straight.  Another  specimen  (PI.  5,  fig.  12)  appears  to  be  proportionately 
longer  but  is  more  compressed.  The  axis,  which  occupies  just  over  one-sixth  of 
the  frontal  breadth,  has  the  outline  of  an  isosceles  triangle,  with  well-defined  straight 
sides  converging  backwards  at  about  20  °.  There  are  at  least  eleven  axial  rings  on 
the  first  specimen,  thirteen  on  the  other,  and  the  axis  ends  in  a  very  small  terminal 
piece  just  short  of  the  posterior  margin.  The  side-lobes  have  an  almost  flat  surface 
which  apparently  becomes  slightly  declined  near  the  pygidial  margin,  and  they 
carry  from  ten  to  twelve  pairs  of  pleural  ribs  in  addition  to  the  pair  of  anterior 
half- ribs.  The  well-defined  pleural  furrows  become  progressively  more  strongly 
directed  backwards  towards  the  rear  of  the  pygidium.  They  are  narrow  and  almost 
straight  over  most  of  their  length  (tr.)  but  within  a  short  distance  of  the  margin 
they  turn  backwards  and  die  out,  apparently  without  attaining  the  margin,  so  that  a 
narrow,  smooth  border  results.  In  a  few  cases  the  furrows  appear  to  intersect  the 
margin,  but  this  may  be  due  to  crushing.  Faint  interpleural  furrows  are  visible  on 
the  adaxial  portions  of  the  first  five  or  so  ribs  but  become  obsolete  on  the  remaining 
ribs. 

Discussion.  Many  of  the  various  species  of  Dionide  have  been  enumerated  and 
discussed  during  recent  years  by  both  Whittington  (1952)  and  Whittard  (1958). 
The  Turkish  specimens  bear  an  overall  resemblance  to  Dionide  formosa  (Barrande) 
and  are  probably  of  broadly  similar  age,  but  there  are  sufficient  minor  differences  to 
warrant  their  separation  as  a  new  subspecies.  First,  the  cranidium  of  D.  formosa 
formosa  has  longer  posterolateral  genal  prolongations,  whilst  the  cephalic  fringe 
contains  a  more  conspicuous  zone  of  marginal  pits,  larger  and  slightly  more  widely- 
spaced  than  those  of  D.  formosa  anatolica.  Second,  although  the  nervures  of  the 
two  forms  are  of  similar  type  those  of  the  Turkish  subspecies  are  apparently  less 
strongly  developed  (judging  from  Whittington's  illustrations,  1952,  pi.  1,  figs.  1,  2, 


112  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

5)  whilst  one  specimen  shows  an  additional  branch.  Third,  the  pygidium  of  D. 
formosa  anatolica  is  better  segmented,  with  a  greater  number  of  both  axial  rings  and 
pleural  furrows.  The  conspicuous  nervures  of  D.  formosa  anatolica  easily  distinguish 
the  new  subspecies  from  forms  such  as  D.  jubata  Raymond  (of  Llanvirn  age)  and 
D.  euglypta  (Angelin)  var.  quadrata  Whittard  (of  Caradoc  age  ;  see  Whittard  1958  : 
99-102).  Dionide  atra  Salter  may  also  be  of  Caradoc  age  but  has  an  unusually  and 
distinctively  large  number  of  axial  rings  and  pleural  ribs  (Whittard  1959  :  98). 
Dionide  turbulli  Whittington  (1952  :  8),  from  the  Llanvirn  Series,  differs  from  D. 
formosa  anatolica  in  having  a  much  broader  cephalic  fringe  with  radiating  ridges, 
very  long  genal  prolongations,  and  a  pair  of  strongly-developed,  single  nervures. 
Dionide  hybrida  Reed  (1915  :  26,  pi.  5,  fig.  7),  from  the  Hwe  Mawng  Beds  of  Burma 
(exact  age  uncertain),  has  a  cephalic  outline  similar  to  that  of  D.  formosa  anatolica 
but  the  glabella,  which  carries  a  median  tubercle,  is  relatively  larger,  whilst  the 
fringe  is  narrower  and  has  coarser  pitting,  especially  marginally  where  there  is  a 
conspicuous  row  of  large  pits.  No  nervures  are  visible  in  Reed's  illustrations  or 
noted  in  his  description,  but  the  species  is  founded  on  only  a  single,  abraded  specimen. 
Dionide  asiatica,  from  the  Ordovician  of  Eastern  Yunnan,  was  founded  by  Kobayashi 
(1940  :  205)  on  the  specimen  figured  by  Mansuy  (1912  :  37,  pi.  6,  figs.  2a,  b)  as  D. 
formosa.  Though  broadly  similar  to  D.  formosa  formosa  and  D.  formosa  anatolica, 
Mansuy's  specimen  may  be  distinguished  by  the  following  features  :  the  nervures 
are  less  strongly  developed  ;  the  frontal  part  of  the  cephalic  fringe  is  much  narrower 
(sag.)  ;  there  is  a  marginal  row  of  conspicuously  large  fringe  pits  ;  the  pygidium  is 
relatively  longer  and  better  segmented  with  twenty-five  axial  rings  and  about 
sixteen  or  seventeen  pairs  of  pleural  ribs  in  addition  to  the  anterior  pair  of  half-ribs. 

Family  DALMANITIDAE  Vogdes,  1890 

Genus  DALMANITINA  Reed,  1905 

Dalmanitina  proaeva  proaeva  (Emmrich) 

(PI.  6,  figs.  1-9,  11-13  ;   PI.  7,  figs.  4,  5) 

1839     Phacops  proaevus  Emmrich  :  25. 

1956     Dalmanitina  proaeva  proaeva  (Emmrich)  Snajdr  :  513,  pi.  4,  figs.   10,   11  ;    pi.  5,  fig.  4. 
Includes  discussion  of  species. 

The  species  and  subspecies  of  Dalmanitina  in  the  Caradoc  and  Ashgill  Series  of 
Bohemia  have  been  revised  by  Snajdr  (1956).  They  comprise,  in  ascending 
stratigraphical  order  :  D.  proaeva  cilinensis  Snajdr,  Drabov  Beds  ;  D.  proaeva 
socialis  (Barrande),  Letna  Beds  ;  D.  proaeva  proaeva  (Emmrich)  [Snajdr  lists  this 
only  from  the  Chlustina  Beds  but  Havlicek  and  others  (1958)  record  it  also  from  the 
underlying  Cernin  Beds]  ;    and  D.  proaeva  grandis  (Barrande),  Kraluv  Dvur  Beds. 

Three  pygidia  of  D.  proaeva  proaeva  from  the  Chlustina  Beds,  figured  by  Snajdr, 
exhibit  a  small  amount  of  variation.  One  (Snajdr  1956,  pi.  4,  fig.  10)  has  ten  axial 
rings  and  seven  and  a  half  pleural  ribs  ;  another  (pi.  4,  fig.  n)  shows  nine  axial  rings 
and  seven  and  a  half  pleural  ribs  ;  whilst  the  third  (pi.  5,  fig.  4)  has  nine  axial  rings 
and  eight  and  a  half  pleural  ribs,  as  far  as  can  be  judged.     There  is  a  strong  overall 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  113 

resemblance  to  the  pygidium  of  D.  proaeva  socialis  but  the  latter,  while  possessing  a 
similar  number  of  pleural  ribs,  has  deep,  relatively  broad  (exsag.)  interpleural 
furrows,  whereas  those  of  D.  proaeva  proaeva,  although  distinct,  are  narrow  and 
shallow. 

Remains  of  Dalmanitina  were  found,  in  varying  concentrations,  throughout  most 
of  the  Bedinan  Formation,  but  although  a  certain  amount  of  variation  was  observed 
it  has  proved  insufficient  to  demonstrate  the  presence  of  more  than  one  form.  The 
hindmost  axial  rings  and  pleural  ribs  of  some  pygidia  are  not  well  defined  but  in 
general  the  range  of  variation  is  limited.  The  number  of  axial  rings  ranges  from 
eight  to  ten  (rarely  traces  of  an  eleventh),  and  of  pleural  ribs,  from  seven  and  a  half 
to  eight  and  a  half  (very  rarely  six,  in  imperfect  material),  figures  which  accord 
well  with  those  for  D.  proaeva  proaeva.  The  specimens  all  exhibit  the  broad  pleural 
furrows  and  narrow  interpleural  furrows  of  the  Czech  species,  and  there  are 
insufficient  grounds  for  separating  the  two.  Many  of  the  Bedinan  pygidia  end  in  a 
terminal  spine  of  varying  length,  usually  short  on  the  internal  mould,  owing  to 
partial  infilling  of  the  internal  cavity,  but  sometimes  considerably  longer  on  the 
external  mould.  Most  of  the  best-preserved  specimens  have  the  spine  about  as  long 
as  the  remainder  of  the  pygidium,  but  one  is  exceptional  in  being  almost  one-and-a- 
half  times  as  long  (PI.  7,  fig.  5).     The  specimens  are  identical  in  all  other  respects. 

The  cephala  of  both  D.  proaeva  proaeva  and  D.  proaeva  socialis  are  virtually 
identical,  and  are  closely  matched  by  the  Turkish  specimens.  Two  hypostomas 
were  collected  (see  PL  6,  figs.  7,  9)  and  although  slightly  compressed  (one  more  so 
than  the  other)  they  resemble  that  of  D.  proaeva  socialis  illustrated  by  Barrande 
(1852,  pi.  26,  fig.  21). 

Localities  and  horizons.  Dalmanitina  proaeva  proaeva  was  found  in  moderate 
numbers  in  the  lower  part  of  the  Bedinan  Formation  as  seen  to  the  east  of  Sosink. 
The  lowest  record  was  from  locality  A .  1,  but  the  species  proved  more  abundant  higher 
in  the  succession  and  was  found  at  A. 2-6,  especially  A. 3.  In  the  Bedinan  district, 
presumably  higher  in  the  succession,  D.  proaeva  proaeva  was  collected  from  many 
of  the  localities  examined  but  occurred  in  greatest  abundance  in  the  upper  part  of  the 
exposed  section,  covered  by  localities  B.  17-21. 

Genus  KLOUCEKIA  Delo,  1935 

Kloucekia  phillipsii  (Barrande)  euroa  subsp.  nov. 

(PI.  6,  fig.  10  ;   PL  7,  figs.  1-3,  6,  7,  9,  12) 

Diagnosis.  Subspecies  of  Kloucekia  phillipsii  dintinguished  principally  by 
the  pygidium,  which  has  eight  or  nine  axial  rings  and  usually  five  or  six  pairs  of 
pleural  ribs,  the  latter  separated  from  each  other  by  deep  pleural  furrows  carrying 
moderately-impressed  interpleural  furrows. 

Holotype.     It.1188  (PL  6,  fig.  10  ;   PL  7,  figs.  9,  12). 

Paratypes.  It.1187  (PL  7,  figs.  6,  7)  ;  It.1189  (PL  7,  fig.  3)  ;  It.  1223  (PL  7, 
fig.  2)  ;   It.  1225  (PL  7,  fig.  1). 

Localities  and  horizons.  K.  phillipsii  euroa  is  one  of  the  most  abundant  and 
characteristic  trilobites  of  the  Bedinan  Formation  in  the  Bedinan  district,  where  it 

GEOL.   15,  2.  12 


n4  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

was  collected  from  all  but  the  lowest  strata.  It  was  not  found  in  the  Bedinan 
Formation  east  of  Sosink,  probably  because  the  beds  there  are  stratigraphically  lower 
than  those  near  Bedinan. 

Description.  The  type  species  of  Kloucekia,  Phacops  phillipsii  Barrande  (1846  : 
27  ;  1852  :  557,  pi.  22,  figs,  i,  2,  pi.  26,  figs.  31-36),  has  recently  been  redescribed  by 
Whittington  (1962  :  7,  text-fig.  2«-i)  using  specimens  named  originally  by  Barrande. 
Most  of  these  came  from  Zahorzany,  Bohemia,  in  strata  once  known  as  the 
Zahorzany  Beds  but  which  would  nowadays  be  termed  the  Chlustina  Beds,  of  fairly 
high  Caradoc  age.  From  a  study  of  Whittington's  illustrations  it  is  clear  that  there 
are  many  points  of  resemblance  between  the  Czech  and  Turkish  specimens,  and  it 
has  not  proved  possible  to  make  a  satisfactory  differentiation  on  the  basis  of  the 
cephalon  and  thorax.  In  certain  cephala  from  the  Bedinan  district  the  glabellar 
furrows  appear  to  run  backwards  from  the  axial  furrows  a  little  more  strongly  than 
do  those  of  K.  phillipsii,  but  this  could  easily  be  accounted  for  by  variation  in 
preservation.  Some  of  the  Turkish  specimens  (see  especially  PI.  7,  fig.  2)  have  a 
V-shaped  group  of  tubercles  on  the  median  portion  of  the  frontal  glabellar  lobe  ; 
such  a  feature,  though  not  uncommon  in  numerous  phacopid  and  dalmanitid  genera, 
is  not  always  preserved.  One  of  Whittington's  photographs  (1962,  text-fig.  2I1) 
shows  that  the  cephalic  doublure  of  K.  phillipsii  possesses  a  vincular  furrow,  and  a 
similar  structure  is  found  in  K.  phillipsii  euroa  (see  PI.  7,  fig.  7),  though  perhaps  a 
little  more  strongly  developed. 

The  most  obvious  differences  between  the  two  forms  are  to  be  found  in  the 
pygidium.  That  of  K.  phillipsii  has  three  well-defined  axial  rings,  followed  by  two 
fainter  rings  and  a  small  terminal  piece,  whilst  the  side-lobes  have  four  pairs  of  deep 
pleural  furrows,  the  ribs  so-formed  carrying  faint  interpleural  furrows.  The 
pygidial  axis  of  K.  phillipsii  euroa  has  five,  occasionally  six,  well-defined  axial  rings, 
followed  by  three  less  well-defined  rings  (the  last  two  ring  furrows  do  not  cross  the 
sagittal  line),  and  ends  in  a  very  small  terminal  piece.  The  side-lobes  of  the  Turkish 
subspecies  usually  carry  five  pairs  of  deep  pleural  furrows,  though  sometimes  there 
are  six  or,  more  rarely,  seven  pairs.  Five  well-developed  pairs  of  interpleural  fur- 
rows are  present,  which  cross  and  indent  the  otherwise  almost  smooth,  gently  concave 
border,  and  there  is  usually  a  trace  of  a  sixth  pair.  In  general  the  pygidium  appears 
to  be  proportionately  broader  than  that  of  K.  phillipsii,  but  it  would  be  unwise  to 
assume  that  this  is  the  case  when  the  material  is  usually  somewhat  compressed. 
According  to  Snajdr  (1956  :  39)  the  vertical  range  of  K.  phillipsii  in  Bohemia  is 
from  the  Drabov  Beds  to  the  Chlustina  Beds  of  the  Caradoc  Series,  but  the  range  of 
K.  phillipsii  euroa  may  be  more  restricted  as  it  has  not  been  found  in  the  lowest  part 
of  the  Bedinan  Formation,  nor  has  it  yet  been  recorded  outside  Turkey. 

Family  CHEIRURIDAE  Salter,  1864 

Cheirurid  gen.  et.  sp.  ind. 

(PI.  10,  fig.  3) 

An  incomplete  thoracic  segment,  figured  here  as  a  latex  cast,  is  the  only  representa- 
tive of  the  cheirurids  so  far  found  in  the  Bedinan  Formation.     The  specimen  shows 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  ti5 

part  of  the  axial  ring,  bearing  a  small  axial  lobe,  delimited  by  an  axial  furrow  which 
is  shallow  medially  and  deepens  to  both  front  and  back.  The  pleura  is  parallel- 
sided  for  most  of  its  length  (tr.)  and  curves  backwards,  at  first  only  gently  and  then 
more  strongly,  to  end  in  a  long,  pointed  tip,  directed  posterolateral^7 .  The  posterior 
margin  has  a  small  posterior  flange,  the  outer  part  of  which  is  not  preserved,  whilst 
an  anterior  flange  of  generally  similar  size  expands  conspicuously  towards  the 
fulcrum,  where  it  ends  in  a  projecting,  articulating  process.  A  shallow  pleural 
furrow  divides  the  pleura  into  two  subequal  bands,  the  anterior  of  which  is  slightly 
the  narrower  (exsag.).  Along  the  pleural  furrow  is  distributed  a  somewhat  irregular 
row  of  almost  equisized  pits  which  begins  just  outside  the  axial  furrow  and  ends 
without  quite  reaching  the  fulcrum.  The  material  is  insufficient  for  firm  identifica- 
tion, but  similar  characters  are  to  be  seen  in  the  thorax  of  Placoparina,  a  genus  of 
Llanvirn-Llandeilo  age  (see  Whittard  1958  :  112). 

Locality  and  horizon.  B.12,  to  the  west  of  Bedinan,  in  the  upper  part  of  the 
Bedinan  Formation  exposed  in  the  section  there. 

Family  SYNHOMALONOTIDAE  Kobayashi,  1963 
Genus  NESEURETUS  Hicks,  1876 
Subgenus  NESEURETINUS  nov. 

Type  species.     N.  (Neseuretinus)  turcicus  sp.  nov. 

Diagnosis.  Subgenus  of  Neseuretus  distinguished  by  large,  inclined,  anterior 
border,  with  convex,  transversely  straight,  preglabellar  field  delimited  by 
conspicuous  preglabellar  and  anterior  border  furrows. 

Distribution.     South-eastern  Turkey,  Burma  and  southern  China. 

Neseuretus  (Neseuretinus)  turcicus  sp.  nov. 
(?P1.  7,  figs.  8,  10,  11;   PI.  9,   figs.  1-4) 

Diagnosis.     As  for  subgenus. 

Holotype.     It.1179  (PL  9,  figs.  1-3). 

Paratype.     It.  1205  (PI.  9,  fig.  4). 

Locality  and  horizon.  The  species  is  known  with  certainty  from  only  locality 
B.2  at  the  section  south-west  of  Bedinan,  where  it  occurs  in  the  lower  part  of  the 
Bedinan  Formation.  A  pygidium  tentatively  referred  to  the  new  form  (see  PI. 
7,  figs.  8,  10,  11)  was  recovered  from  a  probably  similar  horizon  at  locality  A. 3, 
east  of  Sosink. 

Description.  The  species  is  represented  by  two  incomplete  cranidia  preserved 
as  internal  and  external  moulds.  The  cranidium  is  moderately  convex  both 
longitudinally  and  transversely,  with  median  length  estimated  as  being  slightly 
more  than  half  the  maximum  breadth.  The  glabella  is  fairly  convex,  especially 
transversely,  trapezoidal  in  outline,  narrowing  forwards  so  that  the  frontal  breadth 
is  a  little  more  than  half  the  basal  breadth.     There  are  four  pairs  of  unequal 

GEOL.  15,  2.  I2§ 


lib  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

glabellar  lobes,  diminishing  in  size  from  back  to  front  of  the  glabella.  The  fourth 
lobes  are  the  largest,  their  length  about  one-third  that  of  the  glabella,  and  subcir- 
cular  in  outline,  becoming  subangular  anterolaterally.  The  fourth  glabellar  furrows 
are  deeply  incised,  running  straight  inwards  and  slightly  backwards  from  the  axial 
furrows  for  more  than  half  their  length  (tr.)  but  then  turning  more  sharply  backwards 
and  terminating  so  as  to  leave  a  pair  of  "  necks  "  connecting  the  basal  lobes  to  the 
median  body.  There  is  no  obvious  bifurcation  of  the  basal  furrows  but  at  their 
midpoints  they  become  notably  broader,  with  the  adaxial  margins  steeply  bevelled. 
The  third  glabellar  lobes  are  transversely  subrectangular  in  plan,  their  long  axes 
strongly  divergent  forwards,  and  they  are  bounded  by  parallel,  well-defined  third 
glabellar  furrows  which  are  deepest  adaxially.  The  second  glabellar  lobes  are 
generally  similar  to,  though  smaller  than,  the  third  pair,  but  expand  slightly  at  their 
outer  ends,  the  second  glabellar  furrows  becoming  correspondingly  less  divergent 
forwards.  The  second,  third  and  fourth  pairs  of  furrows  end  adaxially  in-line  so  as 
to  leave  a  smooth  median  body  which  is  almost  parallel-sided  and  occupied  just  less 
than  one-third  of  the  basal  glabellar  breadth.  Immediately  in  front  of  the  second 
glabellar  lobes  the  glabella  narrows  slightly,  forming  a  small  "  step  "  in  the  outline. 
The  first  glabellar  lobes  are  very  small,  about  half  the  length  (exsag.)  of  the  first 
pair,  and  poorly  defined  frontally  by  first  glabellar  furrows  which  are  little  more  than 
inconspicuous,  shallow  notches.  The  frontal  glabellar  lobe  is  very  short,  and  the 
almost  straight  anterior  margin  has  a  shallow  median  indentation.  The  length 
(exsag.)  of  the  frontal  lobe  diminishes  markedly  towards  the  posterolateral  angles, 
which  are  less  well  defined  than  the  remainder  of  the  lobe  and  appear  to  extend 
abaxially  just  beyond  the  first  glabellar  lobes.  The  axial  furrows  are  deep  and 
broad,  converging  forwards  in  straight  lines  as  far  as  the  second  glabellar  furrows, 
beyond  which  they  become  more  diffuse,  meeting  the  well-defined  preglabellar 
furrow  and  "  anterior  furrows  "  (see  later)  at  a  pair  of  broad  depressions  in  which  a 
pair  of  hypostomal  pits  is  sited  approximately  opposite  the  mid-point  of  the  frontal 
glabellar  lobe.  The  distance  from  the  front  of  the  glabella  to  the  anterior  margin  of 
the  cranidium  is  estimated  to  be  a  little  more  than  two-thirds  of  the  glabellar  length, 
that  is  to  say  it  is  unusually  long  for  the  genus,  and  is  composed  of  two  distinct  parts, 
an  anterior  border  and  what  appears  to  be  a  true  preglabellar  field.  The  anterior 
border  is  relatively  large,  longest  (sag.)  medially  but  shortening  abaxially,  strongly 
arched  transversely,  and  fairly  steeply  inclined  forwards  to  form  a  scoop-like  front 
to  the  cranidium.  The  preglabellar  field  is  well  developed,  its  surface  strongly 
convex,  standing  slightly  higher  than  the  front  of  the  glabella.  In  plan  it  is  trans- 
versely subrectangular,  defined  posteriorly  by  the  almost  straight  preglabellar 
furrow  and  anteriorly  by  a  deep,  broad  (sag.),  anterior  border  furrow  which  is  well- 
rounded  in  cross-section.  The  outer  ends  of  the  preglabellar  field  are  truncated  by 
broad  (tr.),  moderately-deep  furrows  which  form  forwards  extensions  of  the  axial 
furrows.  Whittard  (i960  :  143)  introduced  the  term  "  anterior  furrows "  for 
apparently  similar  structures  seen  in  other  species  of  Neseuretus,  defining  what  he 
described  as  the  "  anterior  area  ",  believed  by  him  to  represent  the  combined  anterior 
border  and  preglabellar  field.  In  the  new  species  there  can  be  little  doubt  that  a 
true  preglabellar  field  is  present,  so  that  the  furrows  truncating  it  laterally  are 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  117 

equivalent  to  the  posterior  portion  of  Whittard's  anterior  furrows.  The  palpebral 
lobes  are  sited  opposite  the  second  glabellar  lobes  and  the  third  glabellar  furrows. 
They  are  unfurrowed,  strongly  convex  outwards  in  plan,  stand  a  little  lower  than  the 
adjacent  parts  of  the  glabella,  and  have  their  surface  gently  declined  adaxially. 
The  fixigenae  are  convex,  arching  downwards  to  both  front  and  rear  from  the 
vicinity  of  the  palpebral  lobes.  The  gonatoparian  facial  suture  is  of  the  type 
characteristic  for  the  genus.  The  surface  of  the  test  is  incompletely  known,  but 
that  of  the  anterior  border  is  finely  granulated,  whilst  the  glabellar  lobes,  median 
body  and  preglabellar  field  carry  coarser  tubercles. 

The  remainder  of  the  exoskeleton  is  unknown  with  certainty  but  a  single, 
incomplete  pygidium  of  appropriate  type  from  near  Sosink  is  referred  questionably 
to  the  new  species  (PI.  7,  figs.  8,  10,  11).  The  specimen  has  undergone  slight 
compression  but  is  not  unduly  distorted.  The  axis  extends  to,  or  almost  to,  the  tip 
of  the  pygidium  and  its  frontal  breadth  is  just  over  one  third  of  the  total  breadth, 
which  is  measured  just  in  front  of  centre.  In  addition  to  the  articulating  half-ring 
there  are  seven  well-defined  axial  rings,  followed  by  two  less  distinct  rings  (as  far 
as  can  be  judged)  and  a  small  terminal  piece.  The  ring  furrows  all  deepen  abaxially 
and  there  is  a  break  in  the  outline  of  the  axis  behind  the  fifth  axial  ring  ;  as  far  as 
this  point  the  deep  axial  furrows  converge  backwards  at  about  thirty  degrees,  but 
beyond  it,  to  the  blunt  tip,  they  are  subparallel.  The  side-lobes  are  arched-down 
moderately  and  each  carries  five  deep  pleural  furrows  and  a  sixth,  fainter  furrow. 
The  anterior  half-rib  is  deflected  ventrally  to  form  a  facet  of  only  moderate  size. 
Each  of  the  remaining  ribs  carries  an  interpleural  furrow  which  is  only  faint  over  most 
of  its  length  (tr.)  but  then  forms  a  broad  depression  level  with  the  abaxial  ends  of 
the  adjacent  pleural  furrows  and  is  directed  obliquely  backwards  and  outwards  in 
relation  to  them.  Although  the  specimen  is  incomplete  there  appears  to  be  a 
smooth,  narrow  border. 

Discussion.  Numerous  species  of  Neseuretus  have  been  described,  mostly  from 
the  Arenig  and  Llanvirn  Series,  and  many  of  them  have  been  discussed  by  Whittard 
(i960  :  138-151).  Almost  all  can  be  separated  from  the  new  species  by  their 
possession  of  a  so-called  "  anterior  boss  "  in  front  of  the  glabella  and  their  lack  of  a 
discrete  preglabellar  field,  the  only  form  possessing  an  apparently  similar  structure 
being  Neseuretus  birmanicus  (Reed),  a  species  now  assigned  to  N.  (Neseuretinus). 
This  form  was  described  from  the  Upper  Naungkangyi  Beds  of  Burma  as  Calymene 
birmanica  by  Reed  (1906  :  71,  pi.  5,  fig.  27  ;  1915  :  44,  pi.  8,  figs.  1-5),  and  although 
there  is  still  uncertainty  regarding  the  precise  geological  age  it  may  not  be  far  removed 
from  that  of  the  new  Turkish  species.  More  recently  Lu  (1957  :  288,  pi.  154,  figs.  1, 
2)  has  figured  as  Synhomalonatus  (sic)  birmanica  a  cranidium  and  pygidium  from  the 
"  Middle  Ordovician  "  of  Yunnan.  His  illustrations  are  merely  reproductions  of 
two  of  Reed's  figures  of  1915,  but  the  record  is  interesting  in  that  it  extends  the 
known  geographical  range  of  Neseuretinus  into  China.  In  the  present  state  of 
knowledge  of  the  age  of  the  Burmese  faunas  it  is  not  possible  to  say  whether  N. 
(N.)  turcicus  is  younger  than  N.  (N.)  birmanicus,  but  it  may  easily  be  distinguished 
from  the  latter  species  by  the  more  pointed  and  considerably  longer  anterior  border, 
the  more  convex  preglabellar  field,   and  the  slightly  narrower  glabella.     Reed's 


n8  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

paper  (1915)  shows  that  N.  (TV.)  birmanicus  possesses  a  pair  of  small  paraglabellar 
areas,  which  were  not  apparent  in  the  illustrations  to  his  1906  publication  ;  it  is 
not  yet  clear  whether  corresponding  structures  are  present  in  N.  (N.)  turcicus,  but 
the   appropriate   portion   of  the  cranidium  has  not  been   found  well  preserved. 

Family  EOHOMALONOTIDAE  Hupe,  1953 

Genus  BRONGNIARTELLA  Reed,  1905 

Brongniartella  levis  sp.  nov. 

(PL  8,  figs.  2,  3,  6) 

Diagnosis.  Large  species  of  Brongniartella  with  gently  convex  cranidium  and 
scarcely  defined  glabella.  Cranidium  notably  broad  frontally  with  frontal  margin 
only  slightly  convex  forwards.  Three  pairs  unequal  glabellar  lobes  almost  indis- 
tinguishable. Palpebral  lobes  opposite  third  glabellar  furrows  and  some  distance 
from  glabella.     Large,  poorly-defined  paraglabellar  areas  present. 

Holotype.     It.1219. 

Locality  and  horizon.  Locality  B.18  in  the  section  west  of  Bedinan.  The 
horizon  is  in  the  highest  part  of  the  mudstone/shale  succession  of  the  Bedinan 
Formation. 

Description.  The  new  species  is  represented  with  certainty  by  only  a  single 
incomplete,  large  cranidium  of  markedly  depressed  form.  The  median  length  is 
51  mm.,  whilst  the  basal  breadth  must  have  been  of  the  order  of  80  mm.,  although 
the  outer  parts  are  incomplete.  The  glabella  is  subtrapezoidal  in  outline  with  a 
length  of  35  mm.,  narrowing  forwards  from  a  basal  breadth  of  about  36  mm. 
(estimated).  The  anterior  and  lateral  margins  are  only  poorly  defined  by  furrows 
which  are  little  more  than  broad,  shallow  indentations  of  the  test.  Glabellar 
lobation  is  almost  indiscernible,  in  addition  to  being  obscured  by  slight  crushing  of 
the  test,  but  there  are  suggestions  of  three  pairs  of  glabellar  lobes  of  markedly 
unequal  size,  the  basal  pair  being  much  the  largest,  occupying  almost  half  the 
glabellar  length.  The  second  glabellar  lobes  are  rather  less  than  half  the  length 
of  the  basal  pair,  whilst  the  first  pair  are  notably  small,  probably  less  than  half  the 
size  of  the  second  pair.  There  is  a  gentle,  outwards  curvature  of  the  axial  furrows 
opposite  the  second  glabellar  lobes,  which  project  a  little  beyond  the  other  pairs  of 
lobes.  The  anterolateral  angles  of  the  glabella  are  slightly  swollen  dorsally  to  form 
a  pair  of  low,  lobe-like  structures,  between  which  the  otherwise  almost  straight 
anterior  margin  of  the  glabella  is  slightly  indented.  Similar  structures  have  been 
observed  elsewhere  in  Brongniartella,  Neseuretus,  and  other  trilobites  of  the  Calymen- 
acea.  A  slight,  longitudinal,  median  ridging  of  the  centre  of  the  glabella,  though 
probably  exaggerated  by  crushing,  is  undoubtedly  original  in  part,  and  represents  a 
feature  known  from  other  species  of  Brongniartella.  The  preglabellar  field  and  anterior 
border  cannot  be  differentiated  but  are  combined  to  form  a  broad  (sag.),  scoop-like 
structure  which  is  moderately  arched  transversely,  gently  inclined  forwards,  and  has 
its  dorsal  surface  slightly  concave.     The  border  is  separated  by  only  a  pair  of  shallow 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  119 

depressions  from  the  anterior  parts  of  the  fixigenae,  which  are  broad  and  gently 
declined  abaxially.  The  anterior  margin  of  the  cranidium  is  broadly  rounded  in 
outline.  The  occipital  ring  is  subrectangular  in  plan,  its  dorsal  surface  almost  flat, 
whilst  the  occipital  furrow  is  broad  (sag.)  and  only  lightly  impressed,  especially 
medially  and  abaxially  where  it  is  almost  obsolete.  It  curves  forwards  slightly  at 
its  outer  ends  where  the  occipital  ring  passes  almost  imperceptibly  into  what  remains 
of  the  posterior  border.  The  palpebral  lobes  are  set  fairly  well  back,  behind  centre 
of  the  glabella  and  opposite  the  third  glabellar  furrows,  and  stand  only  slightly  lower 
than  the  highest  part  of  the  cranidium.  In  plan  they  are  moderately  convex 
outwards,  particularly  well  defined  at  their  posterior  ends,  and  carry  no  trace  of 
palpebral  furrows.  The  portions  of  the  fixigenae  behind  the  eyes  are  more  steeply 
declined  abaxially  than  are  the  anterior  parts.  They  contain  a  pair  of  roughly 
quadrant-shaped  paraglabellar  areas  which  are  large,  extending  forwards  as  far  as 
the  third  glabellar  furrows,  but  poorly  defined  by  traces  of  shallow  furrows.  The 
facial  suture  is  known  from  only  the  left  anterior  branch.  This  is  sigmoidal  in 
plan,  at  first  curving  forwards  and  slightly  inwards  from  the  eye,  but  then  turning 
gently  outwards  until  it  reaches  the  cephalic  border  longitudinally  almost  in-line 
with  the  eye  ;  finally  it  sweeps  inwards  in  a  broad  curve  to  meet  the  cephalic  margin 
approximately  level  with  the  posterolateral  corner  of  the  glabella.  The  surface  of 
the  test  is  smooth  except  for  the  extreme  edge  of  the  anterior  border,  which  is 
finely  granulated.  In  addition  there  are  widely-spaced,  conspicuous  pits  on  the 
fixigenae  anterior  to  the  eye,  whilst  similar  punctae  are  grouped  more  closely  on  the 
anterior  half  of  the  frontal  cephalic  border,  forming  a  zone  parallel  to  the  anterior 
margin. 

In  addition  to  the  holotype  of  Brongniartella  levis,  trilobite  fragments  assignable 
to  the  genus  were  found,  though  at  other  levels  and  localities,  in  the  Bedinan  Forma- 
tion south-west  of  Bedinan  (see  PI.  8,  figs,  i,  5).  All  represent  smaller  individuals 
than  the  holotype  and  none  is  sufficiently  well  preserved  for  certain  identification. 
An  incomplete  cranidium  (PI.  8,  fig.  1)  may  represent  an  immature  individual  of 
Brongniartella  levis.  A  fragmentary  thorax  and  pygidium  (PI.  8,  fig.  5)  appear 
typical  for  the  genus  but  are  insufficient  for  detailed  comparison. 

Discussion.  Brongniartella  levis  is  one  of  a  group  of  broadly  comparable  species 
of  the  genus  occurring  over  a  wide  area  in  rocks  belonging  to  the  lower  or  middle 
parts  of  the  Caradoc  Series.  In  the  Anglo-Welsh  area  Brongniartella  caradociana 
Dean  (1961  :  349),  from  the  Costonian  Stage  [Nemagraptus  gracilis  Zone],  is  a 
slightly  smaller  species  distinguished  by  having  a  more  convex  glabella,  the  anterior 
half  of  which  becomes  markedly  narrower,  and  eyes  set  farther  forwards  than  in  the 
Turkish  form.  The  well-known  Brongniartella  bisulcata  M'Coy  sp.  (see  Dean 
1961  :  346),  from  a  Caradoc  horizon  roughly  contemporaneous  with  that  of  B. 
levis,  has  the  eyes  a  little  farther  forwards  than  the  latter  species,  the  sides  of  the 
glabella  are  less  convergent  forwards,  and  the  cranidium  is  proportionately  narrower 
frontally.  Brongniartella  platynota  (Dalman)  [=  B.  inexpectata  Barrande  sp., 
see  Kielan  i960  :  116]  is  a  later  form,  apparently  the  last-known  Brongniartella, 
widely  distributed  in  the  Ashgill  Series  of  Central  Europe  and  Scandinavia.  It  is 
easily  distinguished  from  B.  levis  and  other  species  of  the  genus  by  having  the  eyes 


i2o  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

set  well  forwards,  whilst  the  glabellar  outline  narrows  markedly  and  the  front  of 
the  cranidium  is  narrower  and  more  convex  in  plan. 

Genus  PLATYCORYPHE  Foerste,  1919 
Platycoryphe?  sp. 

(PL  8,  fig.  4) 

A  single,  small  cranidium  is  tentatively  assigned  to  Platycoryphe.  The  glabella 
has  a  basal  breadth  about  one  and  a  half  times  the  median  length  (estd),  there  are 
three  unequal  pairs  of  glabellar  lobes,  and  the  outline  narrows  markedly  in  front  of 
the  second  glabellar  furrows.  There  is  a  suggestion  of  a  right  paraglabellar  area  but 
the  eyes  and  the  front  of  the  cranidium  are  not  preserved. 

The  problems  of  separating  Brongniartella  and  Platycoryphe  have  been  discussed 
elsewhere  (Dean  1961  ;  Whittington  1965).  In  general,  the  glabella  of  Platycoryphe 
is  the  more  strongly  segmented,  a  feature  found  in  the  present  specimen.  The 
specimen  is,  however,  preserved  as  an  internal  mould,  in  which  all  cephalic  furrows 
tend  to  appear  deeper,  and  from  the  position  of  the  glabellar  furrows  one  cannot 
exclude  the  possibility  of  its  being  an  immature  example  of  Brongniartella  levis. 
In  the  absence  of  the  anterior  border  it  seems  better  to  refer  the  specimen,  with  some 
doubt,  to  Platycoryphe. 

Locality  and  horizon.  Locality  B.20,  in  the  Bedinan  Formation  west  of 
Bedinan.  The  horizon  is  very  close  to  others  containing  Brongniartella  and  the 
only  other  place  where  the  two  genera  occur  together  is  in  South  Shropshire,  in 
the  lowest  subdivision  of  the  Caradoc  Series  (Dean  1961). 

Family  GOLPOCORYPHIDAE  Hupe,  1953 

Genus  COLPOCORYPHE  Novak  in  Perner,  1918 

Colpocoryphe  sp. 

(PI.  9,  figs.  5,  6-10) 

This  characteristically  Mediterranean  genus  is  represented  in  the  Bedinan  faunas 
by  only  three  specimens,  one  reasonably  complete  and  two  fragmentary  cranidia. 
The  glabella  seems  relatively  narrow  for  the  genus,  with  a  long,  transversely  rectan- 
gular, frontal  glabellar  lobe,  and  the  axial  furrows  converge  forwards  only  gently. 
The  Turkish  species,  which  may  be  new,  is  probably  the  youngest  member  of  the 
genus  yet  recorded.  Colpocoryphe  grandis  Snajdr  sp.  (1956  :  501,  pi.  3,  figs.  1-9), 
from  the  Drabov  and  Letna  Beds  (low  Caradoc  Series)  of  Bohemia,  is  broadly  similar 
and  has  the  eyes  in  a  corresponding  position,  but  differs  in  having  more  convergent 
axial  furrows  and  a  smaller,  shorter  frontal  glabellar  lobe. 

Figured  specimens.  It.  1 183  (PL  9,  fig.  5);  It.  1 197  (PL  9,  figs.  6-8,  10);  It.  1204 
(PL  9,  fig.  9). 

Localities  and  horizons.  One  specimen  is  from  locality  A. 3,  east  of  Sosink, 
whilst  the  others  are  from  B.i  and  B.2,  south-west  of  Bedinan.  All  these  localities 
are  believed  to  be  in  the  lower  part  of  the  Bedinan  Formation,  and  Colpocoryphe 
is  one  of  the  trilobites  common  to  both  sections. 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  121 

Family  ODONTOPLEURIDAE  Burmeister,  1843 

Subfamily  SELENOPELTINAE  Whittington,  1956 

Genus  SELENOPELTIS  Hawle  &  Corda,  1847 

Selenopeltis  inermis  (Beyrich)  angusticeps  subsp.  nov. 

(PL  io,  figs,  i?,  4,  6?,  7,  8) 

Diagnosis.  Subspecies  of  S.  inermis  characterized  by  having  narrower  glabella, 
with  sides  only  slightly  convex  in  plan.  Eyes  set  less  far  apart  than  in  S.  inermis, 
and  palpebral  lobes  less  convex  abaxially  in  plan.  Frontal  glabellar  lobe  has  forked 
appearance  owing  to  median  depression  extending  backwards  as  far  as  second 
glabellar  furrows. 

Holotype.     It.1195  (PL  10,  figs.  4,  7,  8). 

Localities  and  horizons.  The  holotype  is  from  locality  A. 2  in  the  section  east 
of  Sosink,  whilst  a  small  pygidium  probably  attributable  to  the  same  subspecies 
(see  PL  10,  fig.  6)  was  found  nearby,  at  locality  A. 3.  The  only  other  specimen 
belonging,  probably,  to  S.  inermis  angusticeps,  is  a  fragmentary  thorax  (PL  10, 
fig.  1)  from  locality  B.i,  south-west  of  Bedinan.  All  these  localities  are  believed  to 
occur  within  the  lower  part  of  the  Bedinan  Formation. 

Description.  Selenopeltis  inermis  is  a  well-known  species,  widely  distributed  in 
Europe  and  the  Mediterranean  (Tethyan)  Province,  and  its  various  subspecies 
extend  with  only  relatively  small  modifications  from  the  Arenig  to  Ashgill  Series. 
Beyrich  (1846  :  20,  pi.  3,  figs.  ia-c)  first  described  5.  inermis  from  "  Wessela  ", 
Bohemia,  whilst  the  lectotype  of  Selenopeltis  buchi  (Barrande  1846  :  28),  a  species 
considered  by  Whittard  (1961  :  197)  to  be  a  subjective  synonym  of  5.  inermis, 
came  from  the  Chlustina  Beds  (Caradoc  Series)  of  Zahofany,  Bohemia  (see  Snajdr 
1956  :  501).  The  holotype  cranidium  of  S.  inermis  angusticeps  is  broadly  comparable 
with  published  illustrations  of  the  Bohemian  species  but  may  be  distinguished  by  the 
features  listed  in  the  diagnosis.  In  particular  the  glabellar  sides  of  the  Turkish 
form  appear  almost  straight  by  comparison,  whilst  the  distance  from  the  palpebral 
lobe  to  the  axial  line  is  conspicuously  less  than  in  the  Czech  specimens.  The 
fragment  of  thorax  from  east  of  Sosink  shows  no  diagnostic  features  but  the  small 
pygidium  from  near  Bedinan,  although  close  to  that  of  5.  inermis  inermis,  appears  to 
be  slightly  longer  and  has  a  median  indentation  of  the  margin.  However,  this 
specimen  is  too  poorly  preserved  for  detailed  comparison  and  neither  it  nor  the 
thoracic  fragment  is  included  as  type  material.  S.  inermis  angusticeps  is  not  yet 
known  from  anywhere  but  the  Sosink-Bedinan  region  but  it  is  interesting  to  note 
that  Seilacher  (1963  :  530,  fig.  2)  has  recorded  Selenopeltis  buchi  from  the  Sinat  Shales 
in  northernmost  Iraq,  not  far  to  the  east  of  the  Turkish  outcrops. 

Family  ASAPHIDAE  Burmeister,  1843 
Asaphid  gen.  et  sp.  indet. 

(PL  10,  figs.  2,  5) 

A  single  hypostoma,  figured  here  as  an  internal  mould  and  a  latex  cast  of  the 
corresponding  external  mould,  is  the  only  evidence  of  asaphid  trilobites  yet  known 


122  ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY 

from  the  Bedinan  Formation.  The  maximum  breadth,  measured  almost  across  the 
centre,  is  estimated  as  about  three-quarters  of  the  maximum  length.  The  overall 
outline  is  suboval  but  the  posterior  margin  is  bifid,  with  a  large,  parabolic,  median 
indentation  extending  for  about  one  quarter  of  the  length  of  the  entire  hypostoma. 
There  is  a  large,  subcircular,  median  lobe  of  low  convexity,  circumscribed  by  a  furrow 
which  is  deepest  posterolaterally.  The  median  lobe  is  separated  from  the  apex  of 
the  median  indentation  by  a  narrow  {sag.),  flat  strip  equal  to  about  one  tenth  of  the 
total  length  of  the  hypostoma,  and  ending  laterally  in  a  pair  of  poorly-defined 
maculae.  Although  the  front  of  the  hypostoma  is  incomplete,  and  there  is  no  trace 
of  anterior  wings,  there  is  a  flattened  lateral  border,  bearing  terrace-lines  on  its 
dorsal  surface,  which  is  continuous  with  the  bifurcated  posterior  border. 

Figured  specimen.     It.  1203. 

Locality  and  horizon.     Locality  A. 6,  east  of  Sosink,  the  highest  fossiliferous 
horizon  in  this  particular  section. 


VI.   REFERENCES 
Barrande,   J.     1846.     Notice  pre'liminaire  sur  le  Systeme  Silurien  et  les  trilobites  de  Boheme. 
vi  +  91  pp.     Leipzig. 

1852.     Systeme  silurien  du  centre  de  la  Boheme.     Iere  partie.      Recherches  paleontologiques . 

xxx  +  935  pp.,  51  pis.     Prague  &  Paris. 

Beyrich,  E.      1846.      Untersuchungen  iiber  Trilobiten.     37  pp.,  4  pis.     Berlin. 

Boucek,   B.      1937.     Stratigraphie  et  parallelisme  de  l'Ordovicien  superieur  de  la  Boheme. 

Bull.  Soc.  geol.  Fr.,  Paris  (5)  7  :  439-458,  pis.  A-C. 
Cooper,  B.  N.      1953.     Trilobites  from  the  Lower  Champlainian  formations  of  the  Appalachian 

Valley.     Mem.  geol.  Soc.  Am.,  Washington,  55  :  1-69,  pis.  1-19. 
Dean,  W.  T.     1961.     The  Ordovician  trilobite  faunas  of  South  Shropshire,  II.     Bull.  Br.  Mus. 

nat.  Hist.  (Geol.),  London,  5  :  311-358,  pis.  49-55. 

1967.     The  distribution  of  Ordovician  shelly  faunas  in  the  Tethyan  region.     Systematics 

Assoc,  London  [in  press]. 

Emmrich,  H.  F.      1839.     De  Trilobitis.     56  pp.,  1  pi.     Berlin. 

Havlicek,  V.     1950.     The  Ordovician  Brachiopoda  from  Bohemia.     Rozpr.  Ustfed.  Ust.  geol., 

Prague,  13  :  1-133,  pis.  1— 13.     [In  Czech,  with  English  and  Russian  summaries] 
Havlicek,  V.,  Horny,  R.,  Chlupac,  I.  &  Snajdr,  M.      1958.     Fuhrer  zu  den  Geologischen 

Exkursionen  in  das  Barrandium.     Sbir.  geol.  Pritv.,  Prague,  1  :  1-171,  pis.  1-16. 
Kellogg,  H.  E.     i960.     The  geology  of  the  Derik-Mardin  area,  south-eastern  Turkey.     Rept 

Explor.  Divn  Amer.  Overseas  Petr.  Ltd,  Ankara  [unpublished]. 
Kielan,   Z.      i960.     Upper  Ordovician  trilobites  from  Poland  and  some  related  forms  from 

Bohemia  and  Scandinavia.     Palaeont.  pol.,  Warsaw,  11  :  1-198,  pis.  1-36. 
Kobayashi,  T.      1940.     Notes  on  the  Dionideidae.     Jap.  J .  Geol.  Geogr.,  Tokyo,  17  :  203-208. 
Lu,  Y.  H.     1957.     Index  fossils  of  China.     Invertebrate,  PI.  3,  Trilobita.     Acad,  sinica,  Peking  : 

249-294,  pis.  137-155. 
Mansuy,    H.     1912.     Etude    geologique    du    Yun-nan    Oriental.     He    partie,    Paleontologie. 

Mem.  Serv.  geol.  Indoch.,  Hanoi,  1,  2  :  1-146,  pis.  1-25. 
Marek,  L.     1952.     Contribution  to  the  stratigraphy  and  fauna  of  the  uppermost  part  of  the 

Kraluv  Dvur  Shales  (Ashgillian).     Sborntk  tistred.  iJst.  geol.,  Prague,  19  :  429-455,  2  pis. 

[In  Czech,  with  Russian  and  English  summaries.] 
Oehlert,  D.-P.     1895.     Sur  les  Trinucleus  de  l'Ouest  de  la  France.     Bull.  Soc.  geol.  Fr.,  Paris, 

23  :  299-336,  pis.  1,  2. 


ORDOVICIAN  TRILOBITE  FAUNA  OF  S.E.  TURKEY  123 

Reed,  F.  R.  C.     1906.     The  Lower  Palaeozoic  fossils  of  the  Northern  Shan  States,  Burma. 

Palaeont.  indica,  Calcutta  (N.S.),  2,  3  :  1-154,  pis.  1-8. 
1915.     Supplementary  Memoir  on  new  Ordovician  and  Silurian  fossils  from  the  Northern 

Shan  States.     Palaeont.  indica,  Calcutta  (N.S.),  6,  1  :  1-98,  pis.  1-12. 
Schmidt,  G.  C.     1965.     Chart  1.     Proposed  Rock  Unit  Nomenclature,  Petroleum  District  V, 

S.E.  Turkey.     Revised  Edn.     Stratigraphic  Committee,  Turkish  Assoc.  Petroleum  Geologists, 

Ankara,     [unpublished] . 
Seilacher,  A.     1963.     Kaledonischer  Unterbau  der  Irakiden.     Neues  Jb.  Miner.  Geol.  Palaont. 

Mh.,  Stuttgart,  10  :  527-542,  figs.  1-3. 
Snajdr,  M.     1956.     The  trilobites  from  the  Drabov  and  Letna  Beds  of  the  Ordovician  of 

Bohemia.     Sb.  ustred.  tjst.  geol.,  Prague,  22  :  477-533,  pis.  32-37.     [In  Czech,  with  English 

and  Russian  summaries.] 
Tolun,  N.      i960.     Stratigraphy  and  Tectonics  of  southeastern  Anatolia.     Rev.  Fac.  Sci.  Univ. 

Istanbul,  B  25 :  201-264,  2  pls- 
Tolun,  N.   &  Ternek,   Z.     1952.     Notes  geologiques  sur  la  region  de  Mardin.     Turk.  Jeol. 

Kur.  Billt.  Ankara,  3:  1-19,  5  pis. 
Whittard,  W.  F.     1955.     The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire,  I. 

Palaeontogr .  Soc.  [Monogr.],  London  :  1-40,  pis.  1-4. 
1956.     The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire,  II.     Palaeontogr. 

Soc.  [Monogr.],  London  :  41-70,  pis.  5-9. 
1958.     The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire,  III.     Palaeontogr. 

Soc.  [Monogr.],  London  :  71-116,  pis.  10-15. 

1959-     The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire,  IV.     Palaeontogr. 

Soc.  [Monogr.],  London  :  1 17-162,  pis.  16-21. 

1961.     The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire,  V.     Palaeontogr. 

Soc.  [Monogr.],  London  :  163-196,  pis.  22-25. 

1961.     The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire,  VI.     Palaeontogr. 

Soc.  [Monogr.],  London  :  197-228,  pis.  26-33. 

Whittington,  H.  B.     1940.     On  some  Trinucleidae  described  by  Joachim  Barrande.     Amer. 

J.  Sci.,  Washington,  238  :  241-259,  pis.  1-4. 
1950.     Sixteen    Ordovician    genotype    trilobites.     /.    Paleont.,    Menasha,    24  :  531-565, 

pis.  68-75. 

1952.     The  trilobite  family  Dionididae.     J .  Paleont.,  Menasha,  26  :  1— 11,  pis.  1,  2. 

1959-     Silicified  Middle  Ordovician  trilobites  :    Remopleurididae,  Trinucleidae,   Raphio- 

phoridae,  Endymioniidae.     Bull.  Mus.  comp.  Zool.  Harv.,  121  :  369-496,  pis.  1-36. 
1962.     A  Monograph  of  the  Ordovician  trilobites  of  the  Bala  area,  Merioneth,  I.     Palae- 

ontgr.  Soc.  [Monogr.],  London  :  1—32,  pis.  1-8. 

1965.     Platycoryphe,   an   Ordovician  homalonotid   trilobite.     /.   Paleont.,   Menasha,   39  : 

487-491,  pi.  64. 

1966.     Phylogeny  and  distribution  of  Ordovician  trilobites.     /.  Paleont.,  Menasha,  40  : 

696-737,  16  figs. 
Williams,  A.     1948.     The  Lower  Ordovician  Cryptolithids  of  the  Llandeilo  District.     Geol. 

Mag.  Lond.,  85  :  65-88,  pi.  6. 
Yi,  Y.  E.     1957.     The  Caradocian  trilobite  fauna  from  the  Yangtze-Gorges.     Acta  palaeont. 

sinica,  Peking,  5  :  545-559.  pis.  1-5- 


EXPLANATION  OF  PLATES 

All  the  specimens  are  from  the  Bedinan  Formation  and  are  housed  in  the  British 
Museum  (Natural  History),  the  registration  numbers  bearing  the  prefix  It.  They 
are  preserved  as  internal  or  external  moulds,  the  latter  usually  being  figured  in  the 
form  of  latex  casts.  A  light  coating  of  ammonium  chloride  was  applied  before 
photographing.     Plates  i  to  4  by  the  writer  ;    Plates  5  to  10  by  Mr.  P.  J.  Green. 


PLATE  1 

Marrolithoides  orthogonius  sp.  nov.     p.  96 

Bedinan  Formation,  locality  A. 3,  east  of  Sosink. 

Fig.  1.     Latex  cast  of  cranidium.     Paratype,  It. 812,   X5. 

Fig.  2.     Left  half  of  internal  mould  of  small  cranidium.     Paratype,  It. 803,   xg. 

Fig.  3.     Ventral  side  of  internal  mould  of  cephalon  showing  girder.     Paratype,  It. 819,    X4. 

Fig.  5.  Right  side  of  cranidium  showing  anterolateral  angulation  of  outline,  with  small, 
marginal  projection  at  about  R16.     It. 806,   x8. 

Fig.  6.     Internal  mould  of  cranidium.     Holotype,  It. 1200,   X3. 

Fig.  7.  Cranidium  (Meraspis,  Degree  unknown)  showing  only  two  frontal  rows  of  pits, 
increasing  to  three  at  R2  and  four  at  about  R9.  Note  the  alar  lobes  and  eye-ridges.  Paratype, 
It.762,   X15. 

Fig.  9.     Internal  mould  of  dorsal  exoskeleton,  Meraspis  Degree  4.     Paratype,  It. 818,    x8. 

Bedinan  Formation,  locality  A. 6,  east  of  Sosink. 

Fig.  4.  Group  of  three  individuals,  one  enrolled,  preserved  as  internal  moulds.  Paratype, 
It.747,   X4- 

Fig.  8.  Internal  mould  of  cranidium.  Note  also  cranidium  of  small  Meraspis  to  right,  and 
pedicle  valve  of  a  brachiopod,  Aegiromena,  in  top  left  corner.     Paratype,  It. 749,   x6. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  1 


GEOL.  15,  2. 


13 


PLATE  3 

Cryptolithus?  bedinanensis  sp.  nov.  p.  104 

Bedinan  Formation,  locality  B.6,  south-west  of  Bedinan. 

Fig.  1.  Internal  mould  of  cranidium,  It. 663,  X4.  This  is  one  of  the  earliest,  rare  members 
of  the  species  to  have  4  frontal  rows  of  pits. 

Figs.  2,  3.  Oblique  posterolateral  and  plan  views  of  internal  mould  of  dorsal  exoskeleton. 
Holotype,  It.  12 10,    x6,  shows  blunt  pleural  tips.     Fig.  3,   X4. 

Fig.  6.  Latex  cast  of  two  small  individuals,  probably  Meraspis,  Degree  5.  Paratype, 
It.1211,   X5. 

Fig.  9.     Internal  mould  of  incomplete  cranidium.     Paratype,  It.1231,   X5. 

Bedinan  Formation,  locality  B.io,  west  of  Bedinan. 

Fig.  4.     Internal  mould  of  cranidium.     It. 713,   X3. 

Fig.  7.  Internal  mould  of  ventral  side  of  fringe,  showing  almost  equal  development  of  girder 
and  two  pseudogirders.     It. 715,   X3. 

Bedinan  Formation,  locality  B.15,  west  of  Bedinan. 
Fig.  5.     Latex  cast  of  cranidium.     It. 856,   X3. 

Cryptolithus?  sp.  juv.   p.  107 

Bedinan  Formation,  locality  B.8,  south-west  of  Bedinan. 

Fig.  8.  Latex  cast  of  cranidium  of  small  Meraspis  [Degree  unknown],  showing  only  two 
concentric  rows  of  pits  both  frontally  and  laterally.     It. 7 18,    X15. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  3 


PLATE  4 

Cryptolithus?  cf.  bedinanensis  sp.  nov.  p.  107 

Bedinan  Formation,  locality  B.13,  north-west  of  Bedinan. 

Fig.  1.     Pathological  cranidium  in  which  the  regular  pit  arrangement  is  disturbed  in  front  of 
the  right  axial  furrow.     It. 837,   X3. 

Cryptolithus?  bedinanensis  sp.  nov.  p.  104 

Bedinan  Formation,  locality  B.12,  north-west  of  Bedinan. 

Fig.  2.     Latex  cast  of  underside  of  fringe,  showing  principal  girder  and  two  weaker  pseudo- 
girders.     It. 836,   X3. 

Fig.  3.     Internal  mould  of  cranidium.     It. 834,   X4. 

Bedinan  Formation,  locality  B.16,  north-west  of  Bedinan. 
Fig.  4.     Internal  mould  of  cranidium  with  four  rows  of  pits  developed  frontally.     It. 867,  X3. 

Bedinan  Formation,  locality  B.22,  south-east  of  Bedinan. 
Fig.  5.     Latex  cast  of  small  cranidium.     It. 890,   x6. 

Bedinan  Formation,  locality  B.13,  north-west  of  Bedinan. 

Fig.  6.  Internal  mould  of  incomplete  cranidium.     It.1212,   X3. 

Fig.  7.  Internal  mould  of  cranidium.     It.  12 16,   X3. 

Fig.  8.  Internal  mould  of  cranidium  with  two  attached  thoracic  segments.     It.1215,    X3. 

Fig.  9.  Internal  mould  of  small  cranidium  showing  reticulation  of  the  cheek-lobes.     It.1214, 

Marrolithoides  sp.  p.  101 

Bedinan  Formation,  locality  B.21,  north-west  of  Bedinan. 

Fig.  10.     Internal  mould  of  cranidium.     Note  long  occipital  spine  and  anterolateral  angulation 
of  cephalic  outline.     It. 881,   X3. 


Bull.  Br.  Mus.  nal.  Hist.  (Geol.)  15,  2 


PLATE  4 


PLATE  5 

Dionide  formosa  (Barrande)  anatolica  subsp.  nov.  p.  109 

Bedinan  Formation,  locality  A. 3,  east  of  Sosink. 

Figs.  1,2.  Plan  and  left  oblique  views  of  cranidium  showing  long  glabellar  spine.  Holotype, 
It.1061,   X4. 

Fig.  3.  Plan  view  of  specimen  showing  left  librigenal  spine  and  part  of  thorax.  Note 
macropleural  first  segment.     Paratype,  It.  1 201,   X5. 

Fig.  12.  Internal  mould  of  pygidium  illustrating  hindmost  pleural  ribs.  Paratype,  It. 1057, 
X4. 

Bedinan  Formation,  locality  A. 6,  east  of  Sosink. 
Fig.  4.     Latex  cast  of  cranidium.     Paratype,  It.  11 78,   X7. 

Bedinan  Formation,  locality  A. 5,  east  of  Sosink. 

Fig.  6.     Internal  mould  of  cranidium.     Paratype,  It. 1062,   x6. 

Fig.  7.  Internal  mould  of  pygidium  with  thorax  lacking  first  segment.  Paratype,  It.  1063, 
X6. 

Atnpyx  nitidus  sp.  nov.  p.  93 

Bedinan  Formation,  locality  B.2,  south-west  of  Bedinan. 

Fig.  5.     Internal  mould  of  pygidium.     Paratype,  It.  1207,   X4. 

Fig.  8.     Internal  mould  of  cranidium.     Holotype,  It. 1181,   X3. 

Fig.  9.     Fragment  of  right  cheek  showing  position  of  eye.     Paratype,  It.  1208,   X4. 

Fig.  10.     Internal  mould  of  small  hypostoma.     Paratype,  It.  11 80,   X12. 

Fig.  11.     Incomplete  cranidium  showing  faint  glabellar  lobes.     Paratype,  It.  1209,   X3. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  5 


PLATE  6 
Dalmanitina  proaeva  proaeva  (Emmrich)  p.  112 
Bedinan  Formation,  locality  B.21,  west  of  Bedinan. 

Fig.  1.     Internal  mould  of  cephalon.     It.  1224,   Xi-5- 

Figs.  4,  11.     Latex  cast  and  internal  mould  of  pygidium.     It.i226,   X2. 

Bedinan  Formation,  locality  B.13,  west  of  Bedinan. 
Fig.  2.     Plan  view  of  pygidium.     It.1190,   X3. 

Bedinan  Formation,  locality  B.20,  west  of  Bedinan. 
Figs.  3,  5.     Internal  mould  of  pygidium.     It. 1222,   X2. 

Bedinan  Formation,  locality  B.3,  west  of  Bedinan. 

Fig.  6.     Internal  mould  of  slightly  disarticulated  specimen.     It. 1 186,   X2-5. 
Fig.  7.     Internal  mould  of  small  hypostoma.     It.1185,   X3. 

Bedinan  Formation,  locality  B.19,  west  of  Bedinan. 

Figs.  8    12.     Plan  and  posterior  views  of  internal  mould  of  pygidium.     It.1193,    xi.     (See 
also  PI.  7,  fig.  5.) 

Bedinan  Formation,  locality  B.2,  west  of  Bedinan. 
Fig.  9.     Internal  mould  of  slightly  compressed  hypostoma.     It.1182,   X2. 

Bedinan  Formation,  locality  B.22,  west  of  Bedinan. 
Fig.  13.     Internal  mould  of  cephalon  and  pygidium  with  long  terminal  spine.     It.1233,  xi-j. 

Kloucekia  phillipsii  (Barrande)  euroa  subsp.  nov.  p.  113 

Bedinan  Formation,  locality  B.13,  south-west  of  Bedinan. 

Fig.  10.     Plan  view  of  internal  mould  of  pygidium  with  thorax  attached.     Holotype,  It. 1 188, 
X3.     (See  also  PI.  7,  figs.  9,  12.) 


Bull.  Br.  Mus.  nal.  Hist.  (Geol.)  15,  2 


PLATE  6 


PLATE  7 

Kloucekia  phillipsii  (Barrande)  euroa  subsp.  nov.  p.  113 

Bedinan  Formation,  locality  B.21,  west  of  Bedinan. 

Fig.  1.     Internal  mould  of  incomplete  cephalon.     Paratype,  It.  1225,   X2. 
Fig.  2.     Plan  view  of  internal  mould  of  cephalon.     Paratype,  It. 1223,   X4. 

Bedinan  Formation,  locality  B.13,  south-west  of  Bedinan. 

Fig.  3.     Internal  mould  of  almost  complete  dorsal  exoskeleton.     Paratype,  It. 1 189,    Xi'5- 
Figs.  6,  7.     Plan  and  oblique  ventral  views  of  cephalon.     Fig.  7  shows  doublure  and  vincular 
furrow.     Paratype,  It.  1 187,   X3. 

Figs.  9,  12.     Thorax  with  attached  pygidium  (see  also  PI.  6,  fig.  10).     Holotype,  It.1188,  X2. 

Dalmanitina  proaeva  proaeva  (Emmrich)  p.  112 
Bedinan  Formation,  locality  B.3,  south-west  of  Bedinan. 
Fig.  4.     Latex  cast  of  cranidium  of  Meraspis  (Degree  unknown).     It.1018,   X12. 

Bedinan  Formation,  locality  B.19,  west  of  Bedinan. 

Fig.    5.     Latex   cast  of  pygidium   with   exceptionally   long   terminal  spine.     It.1193,     xi. 
(See  also  PI.  6,  figs.  8,  12.) 

?  Neseuretus  (Neseuretinus)  turcicus  subgen.  et  sp.  nov.  p.  117 

Bedinan  Formation,  locality  A. 3,  east  of  Sosink. 

Figs.   8,    10,   11.     Posterior,  plan  and  right  lateral  views  of  internal  mould  of  pygidium. 
It.1196.   X2. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  7 


PLATE  8 
Brongniartella  sp.  p.  119 
Bedinan  Formation,  locality  B.13,  south-west  of  Bedinan. 
Fig.  1.     Internal  mould  of  damaged  cranidium.     It.1217,   Xi-5- 

Bedinan  Formation,  locality  B.19,  south-west  of  Bedinan. 
Fig.  5.     Internal  mould  of  thorax  and  pygidium.     It. 1220,   X2-5- 

Brongniartella  levis  sp.  nov.  p.  118 
Bedinan  Formation,  locality  B.18,  south-west  of  Bedinan. 
Figs.  2,  3,  6.     Anterior,  left  lateral  and  plan  views  of  incomplete,  large  cranidium.     Holotype. 
It.1219,   XI-25. 

Platycoryphe  ?  sp.  p.  120 
Bedinan  Formation,  locality  B.20,  south-west  of  Bedinan. 
Fig.  4.     Internal  mould  of  incomplete  cranidium.     It.  1 221,   X3. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  8 


PLATE  9 

Neseuretus  (Neseuretinus)  turcicus  subgen.  et  sp.  nov.  p.  115 

Bedinan  Formation,  locality  B.2,  south-west  of  Bedinan. 

Figs.  1-3.  Plan,  left  lateral  and  oblique  anterolateral  views  of  latex  cast  of  incomplete 
cranidium,  showing  convex  preglabellar  field  and  large,  inclined,  anterior  border.  Holotype, 
It.1179,   X3. 

Fig.  4.  Plan  view  of  internal  mould  of  incomplete  cranidium  lacking  marginal  portion  of 
anterior  border.     Paratype,  It.1205,  X2. 

Colpocoryphe  sp.  p.  120 
Bedinan  Formation,  locality  B.2,  south-west  of  Bedinan. 
Fig.  5.     Latex  cast  of  frontal  portion  of  fragmentary  cranidium.     It.  1 183,   X2. 

Bedinan  Formation,  locality  A. 3,  east  of  Sosink. 

Figs.  6,  10,  latex  cast  and  figs.  7,  8,  corresponding  internal  mould,  of  small  cranidium. 
It.1197,   x6. 

Bedinan  Formation,  locality  B.I,  south-west  of  Bedinan. 
Fig.  9.     Posterior  half  of  glabella,  an  internal  mould.     It.  1204,   X2-5- 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  9 


im  mm 

Si 

10 


PLATE  10 
Selenopeltis  inermis  (Beyrich)  angusticeps  subsp.  nov.  p.  121 
Bedinan  Formation,  locality  B.i,  south-west  of  Bedinan. 
?  Fig.  1.     Internal  mould  of  fragment  of  thorax.     It.  1206,  X2-5. 

Bedinan  Formation,  locality  A. 2,  east  of  Sosink. 
Figs.  4,  7,  8.     Left  lateral,  anterior  and  plan  views  of  internal  mould  of  cranidium.     Holotype, 
It.1195,   X3. 

Bedinan  Formation,  locality  A. 3,  east  of  Sosink. 
?  Fig.  6.     Internal  mould  of  small  pygidium.     It.  1 198,   X5- 

Asaphid  gen.  et  sp.  ind.  p.  121 

Bedinan  Formation,  locality  A. 6,  east  of  Sosink. 

Figs.  2,  5.     Latex  cast  and  internal  mould  of  hypostoma,  showing  indented  posterior  margin. 
It.1203,   X2-5- 

Cheirurid  gen.  et  sp.  ind.  p.  114 

Bedinan  Formation,  locality  B.12,  south-west  of  Bedinan. 

Fig.  3.     Latex  cast  of  left  half  of  thoracic  segment,  showing  furrow  with  pits,  and  anterior 
flange  terminating  in  articulating  process.     It.1213,   X2-5. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  2 


PLATE  10 


J2  JUHW 


PRINTED  IN  GREAT  BRITAIN 
BY  ADLARD  &  SON  LIMITED 
BARTHOLOMEW    PRESS,    DORKING 


BURROWS  AND  SURFACE  TRACES 

FROM  THE  LOWER  CHALK  OF 

SOUTHERN  ENGLAND 


W.  J.  KENNEDY 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  3 

LONDON:  1967 


BURROWS  AND    SURFACE  TRACES  FROM  THE 
LOWER  CHALK  OF  SOUTHERN   ENGLAND 


BY 


WILLIAM  JAMES  KENNEDY 


^ 


Pp.  125-167  ;  9  Plates  ;  7  Text-figures 


BULLETIN  OF 

THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 

GEOLOGY  Vol.  15  No.  3 

LONDON:  1967 


THE    BULLETIN    OF    THE    BRITISH    MUSEUM 

(natural  history),  instituted  in  1949,  is 
issued  in  five  series  corresponding  to  the  Departments 
of  the  Museum,  and  an  Historical  series. 

Parts  will  appear  at  irregular  intervals  as  they  become 
ready.  Volumes  will  contain  about  three  or  four 
hundred  pages,  and  will  not  necessarily  be  completed 
within  one  calendar  year. 

In  1965  a  separate  supplementary  series  of  longer 
papers  was  instituted,  numbered  serially  for  each 
Department. 

This  paper  is  Vol.  15,  No.  3  of  the  Geological 
(Palaeontological)  series.  The  abbreviated  titles  of 
periodicals  cited  follow  those  of  the  World  List  of 
Scientific  Periodicals. 


World  List  abbreviation  : 
Bull.  Br.  Mus.  nat.  Hist.  (Geol.). 


Trustees  of  the  British  Museum  (Natural  History)  1967 


TRUSTEES    OF 
THE    BRITISH    MUSEUM    (NATURAL    HISTORY) 

Issued  24  November,  1967  Price  £2    2s. 


BURROWS  AND   SURFACE   TRACES  FROM  THE 
LOWER  CHALK  OF  SOUTHERN  ENGLAND 


By  WILLIAM  JAMES  KENNEDY 

CONTENTS 


I.  Introduction 

II.  The  trace  fossils 

III.  Location  of  specimens  . 

IV.  Systematic  descriptions 
V.  Other  burrows 

VI.  Conclusions 

VII.  Acknowledgments 

VIII.  References  . 


127 
130 
131 
131 
156 
161 
162 
162 


SYNOPSIS 

The  Lower  Chalk  of  Southern  England  comprises  between  17  and  80  m.  of  rhythmic  alterna- 
tions of  more  or  less  calcareous  marly  chalk.  The  whole  sequence  is  intensely  burrowed,  the 
following  burrow  types  are  named:  Thalassinoides  saxonicus  (Geinitz),  T.  cf.  suevicus  (Rieth), 
T.  ornatus  ichnosp.  nov.,  T.  paradoxica  (Woodward),  Spongeliomorpha  sp.,  Spongeliomorpha? 
annulatum  ichnosp.  nov.  and  Chondrites  sp.  Several  other  forms  are  discussed,  including  "  Tere- 
bella  "  cancellata  Bather  and  Keckia  (?)  sp.  Five  burrow  types  too  poor  for  detailed  description 
are  noted  and  discussed.  Two  other  trace-fossils,  "  laminated  structures  ",  regarded  as  related 
to  T.  saxonicus  and  Pseudobilobites  jefferiesi  ichnosp.  nov.  (the  "  problematicum  "  of  Jefferies 
(1961,  1963))  are  also  described. 

The  Eocene  form  "  Terebella  "  harefieldensis  White  is  briefly  discussed  and  interpreted  as  a 
crustacean  boring.  Thalassinoides ,  Spongeliomorpha  "  laminated  structures  ",  Pseudobilobites, 
"  Terebella  "  cancellata  and  two  of  the  un-named  burrows  are  regarded  as  the  product  of  crusta- 
ceans. Chondrites  and  the  other  un-named  burrows  are  attributed  to  "  worms  ". 

Other  trace  fossils — borings  and  micro-coprolites-are  also  noted. 


I.     INTRODUCTION 

(a)  General  Features.  The  Lower  Chalk  of  Southern  England  comprises  between 
17  and  80  m.  of  marly  blue  or  buff  chalk,  ranging  in  age  from  Lower  to  Upper  Ceno- 
manian.  The  present  account  is  based  chiefly  on  exposures  along  the  North  and 
South  Downs,  the  Chilterns  and  the  Isle  of  Wight,  although  sections  to  the  north, 
at  Hunstanton  (Norfolk)  and  south-west  (Dorset,  Somerset  and  Devon)  have  also 
been  examined. 

Apart  from  the  valuable  coastal  exposures,  there  are  useful  working  sections  around 
Lewes  (Sussex),  Burham  and  Holborough  in  the  Medway  Valley  (Kent)  and  in  the 
large  pits  at  Barrington  (Cambridgeshire),  Houghton  Regis  and  Totternhoe  (Bed- 
fordshire), Pitstone  (Buckinghamshire),  Chinnor  and  Childrey  (Oxfordshire). 

By  far  the  best  general  account  of  the  formation  is  still  that  given  by  Jukes-Browne 
(1903),  whilst  a  brief  outline  of  some  of  the  depositional  and  post-depositional 
features  has  been  given  elsewhere  (Kennedy,  1967). 


GEOL.  15,  3. 


14 


128  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

In  the  Weald,  Hampshire  and  the  Chilterns,  the  base  of  the  formation  is  generally 
marked  by  a  sharp  lithological  break  and  a  line  of  piping,  the  basal  Glauconitic  Marl 
being  let  down  into  the  Gault  or  Upper  Greensand  below,  generally  without  signs  of 
marked  erosion,  and  with  occasional  indications  of  continuous  deposition  from  the 
Albian  below.  The  basal  few  feet,  rich  in  glauconite  and  phosphates,  clearly  indi- 
cates slow  deposition;  similar  features  appear  occasionally  in  the  lower  part  of  the 
Chalk  Marl  above,  as  at  Eastbourne  (Sussex)  and  in  the  Isle  of  Wight.  Above, 
the  whole  thickness  can  be  interpreted  as  a  sequence  of  rhythmic  alternations  of  more 
and  less  calcareous  (or  marly)  chalks,  with  carbonate  contents  varying  between 
approximately  40  %  at  the  base,  increasing  upwards  to  90-95  %  at  the  top  of  the 
Lower  Chalk.  Insoluble  residue  determinations  suggest  that  these  rhythms  are 
visible  when  the  difference  in  carbonate  content  is  as  low  as  4-5  % . 

There  is  much  variation  in  thickness  of  the  "  limestone  "  and  "  marl  "  members 
of  rhythms,  although  in  a  general  way,  in  the  lower  part  the  "  marls  "  are  45-60  cm. 
thick,  the  "  limestones  "  15-30  cm.  thick.  In  the  middle  part  the  alternations  are 
15-30  cm.  thick,  whilst  in  the  upper  part,  there  is  great  variation,  made  difficult  to 
interpret  by  the  low  mud  content  which  renders  the  alternations  only  faintly  discern- 
ible. A  general,  although  not  invariable  feature  of  these  rhythms  is  that  the  "  marl  " 
to  "  limestone  "  contact  is  transitional,  whilst  the  contact  at  the  base  of  the  marls  is 
very  sharp. 

The  following  features  indicate  that  the  alternations  are,  at  least  in  part,  primary : 

(i)  The  piping  of  "  marls  "  into  "  limestones  "  and  vice-versa,  in  a  wide  variety 
of  burrows. 

(ii)  The  occurrence  of  "  limestone  "  pebbles,  phosphatized,  glauconitized  and 
otherwise,  in  "  marls  ". 

(iii)  The  cutting  of  "  marl "-"  limestone  "  junctions  by  erosion  hollows. 

Evidence  of  secondary  segregation  is  suggested  by  the  nodular  appearance  of 
some  "  limestones  "  and  the  occurrence  of  calcareous  concretions  in  some  of  the 
"  marls  ".  In  addition,  sponges,  ammonites  and  other  fossils  in  "  limestones  "  are 
often  undistorted  whilst  the  same  forms  are  crushed  flat  in  "  marls  ",  suggesting  the 
pre-compactional  deposition  of  carbonate  in  the  more  calcareous  parts  of  rhythms. 

The  upper  limit  of  the  Lower  Chalk  in  this  region  is  marked  by  a  sharp  change  in 
lithology  at  the  base  of  the  plenus  Marls,  associated  in  some  areas  with  obvious  signs 
of  erosion,  the  sub-plenus  erosion  surface  of  Jefferies  (1962,  1963). 

Traced  northwards,  the  Lower  Chalk  loses  these  features,  thins  considerably  and 
in  Norfolk  at  Hunstanton  is  clearly  condensed,  with  signs  of  erosion  at  many  levels. 
It  rests,  with  a  sharp  break  and  obvious  signs  of  erosion  on  the  Red  Chalk  (Albian). 
The  Chalk  here  is  hard,  and  as  pointed  out  by  Peake  &  Hancock  (1961)  probably 
winnowed.  These  features,  and  a  similar  thinning  and  condensation  in  the  under- 
lying Albian  suggest  the  presence  of  a  stable  massif  in  this  region  during  part,  at  least, 
of  the  Cretaceous.  The  influence  of  this  massif  may,  in  part  be  responsible  for  the 
development,  in  the  Chilterns  and  northwards,  of  the  "  gritty  "  phosphatic  Tottern- 
hoe  Stone  (Middle  Cenomanian). 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  129 

Traced  westwards,  the  Lower  Chalk  maintains  its  general  features  to  the  western 
limits  of  outcrop  at  Membury  (S.  Devon),  where  rhythms  are  still  present  in  the 
chalky  part  of  the  sequence.  Faunal  evidence  indicates  that  the  base  of  the  Chalk  is 
diachronous  in  the  south-west. 

(b)  The  Chalk.  The  general  composition  of  the  Chalk  was  first  noted  by  Ehren- 
berg  (Sorby  1861)  and  later  by  Sorby  (1861),  Hume  (1893)  and  Jukes-Browne  & 
Hill  (1903,  1904).  More  recently  Black  (1953),  Black  &  Barnes  (1959)  and  Hancock 
(1963)  have  given  additional  information.  The  carbonate  portion  of  the  Chalk  is 
now  wholly  calcite,  and  it  is  generally  accepted  that  most  was  deposited  as  such,  and 
that  it  is  wholly  biogenic  in  origin.  The  finer  fractions  are  largely  composed  of 
coccoliths,  both  whole  and  fragmentary,  whilst  the  coarser  fractions  consist  of  Oligo- 
stegina,  foraminifera,  sponge  fragments,  the  broken-down  prismatic  layers  of 
Inoceramus  and  echinoderm  debris.  Abundance  of  the  latter  gives  rise  to  the 
"  gritty  "  chalks  such  as  the  Totternhoe  Stone  and  Melbourn  Rock.  The  insoluble 
fraction,  discussed  previously  by  Hume  (1893)  and  Hill  (1903,  1904)  includes,  in 
addition  to  the  clay  fraction,  detrital  silt  and  sand  grade  quartz  as  the  most  obvious 
mineral,  accompanied  by  authigenic  glauconite,  collophane  and  feldspar. 

(c)  Bottom  Conditions.  Current  activity  is  indicated  by  the  presence  at  many 
levels  of  winnowed  chalks  and  rolled,  glauconitized  and  phosphatized  pebbles  and 
fossils.  The  body  chambers  of  large  ammonites  are  often  full  of  small  fossils,  includ- 
ing ammonites  up  to  10  cm.  in  diameter,  presumably  swept  in  by  bottom  currents. 
Fragmentation  of  Inoceramus  shells  and  echinoid  tests  may  be  due  to  current  activity. 

Intraformational  conglomerates  suggest  local  erosion,  as  do  what  appear  to  be 
large  scour  hollows,  sometimes  associated  with  large  ammonites  (Kennedy,  1967). 

The  presence  of  burrowing  bivalves  such  as  Pholadomya,  Cucullaea,  and  Panopea 
suggest  soft  bottoms,  as  does  the  presence  of  Teredina  amphisbaena  (Goldfuss),  a 
form  which  I  have  never  seen  associated  with  wood  (although  Teredo  bored  wood 
occurs).  Like  the  recent  Teredo  [Fur cello)  polythalamia  (Linne)  (Oosting  1925)  this 
form  appears  to  have  lived  in  mud.  Although  soft,  the  bottoms  must  have  been  in 
the  form  of  a  stiff  mud,  since  the  small  solitary  corals  and  serpulids  which  are  so 
common  would  not  survive  in  a  fluid  mud,  nor  would  larger  epifaunal  forms  such  as 
the  limid  and  pectinid  bivalves  and  Inoceramus,  the  latter  possibly  byssally  attached 
to  the  sea  floor.  Equally,  the  lobster-like  crustacean  Enoploclytia  presumably 
needed  a  firm  bottom  to  walk  across.  Intense  burrowing  suggests  bottoms  rich  in 
organic  debris. 

There  is  little  evidence  of  rock  bottoms  (hardgrounds)  in  the  Lower  Chalk,  erosion 
surfaces,  when  they  occur,  lacking  the  epifauna  of  bryozoa,  serpulids  and  cemented 
bivalves  present  on  the  Chalk  Rock  hardgrounds.  Borings  in,  and  epifaunas  on, 
the  phosphatized  top  of  the  Upper  Greensand  in  the  south-west  indicate  hard  bottoms 
here  at  least. 

(d)  Depth  of  Deposition.  The  abundance  of  coccoliths  suggests  deposition 
below  the  upper  limit  of  present  day  coccolith  abundance  (60  m.) :  study  of  the 
sponges  indicates  a  depth  of  280  m.  (Cayeux  1897:  Turonian-Senonian)  or  300  m. 
(Gignoux  1926:   Campanian).     Since  it  is  generally  agreed  that  the  Chalk  Marl,  like 


i3o  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

the  Chalk  Rock  (U.  Turonian)  represents  shallower  conditions  than  the  rest  of  the 
Chalk  (Jukes-Browne  &  Hill  1904),  a  depth  lower  than  the  maximum  is  implied  for 
this  part  of  the  Lower  Chalk.  Burnaby  (1962)  has  discussed  depth  variation  in  the 
Lower  Chalk  sea,  on  the  basis  of  the  foraminifera. 


II.     THE  TRACE  FOSSILS 

Two  types  of  trace  fossil  are  described  from  the  Lower  Chalk;  burrows  and  surface 
traces.  Of  these,  burrows  are  by  far  the  most  important,  and  are  one  of  the  most 
prominent  features  of  the  sediment  (Pis.  i,  2). 

(a)  Burrows.  In  modern  marine  environments  a  great  variety  of  organisms 
utilize  the  region  below  the  sediment-water  interface  for  refuge,  nourishment  and 
habitation.  Arthropods,  echinoderms,  molluscs,  coelenterates,  many  groups  of 
worms  (particularly  annelids)  are  amongst  the  most  important  groups  of  inverte- 
brates, whilst  many  higher  animals  burrow.  In  addition,  the  interstitial  fluid  be- 
tween sedimentary  particles  supports  a  large  fauna  and  flora  (Purdy ,  1964) .  Particles 
of  sediment  themselves  have  a  coating  of  bacteria,  utilized  by  detritus  feeders;  total 
content  increasing  as  particle  sizes  decrease  (Newell  1965). 

The  influence  of  these  organisms  on  the  sediment  is  considerable.  Davidson  (1891) 
described  the  activities  of  lobworms  in  the  Holy  Island  Sands,  between  Holy  Island 
and  the  Northumberland  coast,  suggesting  nearly  two  thousand  tons  of  sand  was 
ingested  per  acre  per  annum,  and  that  the  top  60  cm.  of  sediment  passed  through  the 
worms'  bodies  every  22  months.  Taylor  (1964)  quotes  data  suggesting  80-90% 
of  the  sands  in  the  Bermudas  is  made  up  of  ground  shell  matter  that  has  passed 
through  the  intestinal  tracts  of  echinoderms.  Both  indicate  the  importance  of 
biological  destruction  of  sedimentary  structures.  Local  topography  can  be  influenced 
by  burrowing  organisms;  the  hummocky  bottom  topography  of  the  Bahaman 
platforms  is  attributed  to  organic  activity  (Taylor  1964),  whilst  erosion  of  callianassid 
burrows  produces  the  characteristic  sand-pipe  topography  of  some  intertidal  regions 
(Weimer  &  Hoyt  1964).  Many  of  the  problematic  mounds  and  depressions  seen  in 
deep-sea  photographs  are  probably  organic  in  origin. 

Many  burrows  are  lined  with  mucus,  whilst  Callianassa  major  Say  lines  its  burrow 
with  collophane-cemented  sand  pellets  (Weimer  &  Hoyt  1964).  Many  sediment 
eaters  form  durable  faecal  pellets  (Moore  1939).  These  features  indicate  the  im- 
portance of  burrowing  organisms  in  stabilization  and  aggregation  of  sediments. 

Taylor  (1964)  has  pointed  out  the  chemical  effects  of  bottom  dwelling  organisms  on 
both  Eh  and  pH,  particularly  where  the  release  of  organic  and  inorganic  acids  is 
concerned,  suggesting  great  importance  in  diagenesis  at  the  early  burial  stage. 

Burro wers  are  also  responsible  for  the  creation  of  refuges  for  many  commensals. 
The  burrow  of  the  worm  Urechis,  for  instance,  is  inhabited  by  a  gobie,  polynoid  worm 
and  pinnotherid  crab  (Fisher  &  MacGinitie  1928).  Dales  (1957)  gives  details  of 
similar  associations  in  other  burrowing  organisms. 

(b)  Burrows  in  the  Lower  Chalk.  The  whole  of  the  Lower  Chalk  studied  is 
intensely  burrowed  (PL  2,  figs.  2-4),  often  many  times  over  (PL  2,  fig.  4).  In  general, 
these  structures  can  be  studied  in  section  only,  in  the  form  of  sedimentary  mottling. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  131 

Only  rarely  can  the  pattern  and  form  of  systems  be  made  out.  Burrowing,  often 
equally  intense  to  that  in  the  Lower  Chalk,  can  be  seen,  in  suitable  lithologies,  in  the 
overlying  Middle  and  Upper  Chalk. 

Simpson  (1957),  Hantzschel  (1962)  and  Seilacher  (1964)  have  discussed  the  various 
conditions  of  preservation  of  trace  fossils ;  in  the  Lower  Chalk  the  following  modes  of 
preservation  of  burrows  can  be  distinguished : 

(i)  Differences  in  composition  and  colour  of  burrow  filling  and  matrix, 
(ii)  Pyritization  (an  example  of  a  pyritized  burrow,  overgrown  by  a  pyrite  nodule 

was  figured  by  Mantell  (1822,  pi.  16,  fig.  16)). 
(iii)  Coating  of  the  outer  surface  of  the  burrow  by  iron  sulphide  (often  altered  to 

Limonite),  perhaps  influenced  by  the  former  presence  of  a  mucus  lining. 

Burrowing  in  the  English  chalk  has  been  mentioned  only  briefly  by  previous 
authors,  generally  as  "  mottling  "  or  "  piping  ",  or  by  reference  to  them  as  sponges 
(Webster  1814)  or  "Zoophytes"  (Taylor  1823).  More  recently,  Wood  (1965), 
discussing  the  Lower  Chalk  at  Dover,  mentions  "  extensive  reworking  "  by  "  bottom 
living  organisms,  the  infilling  of  the  burrow  traces  being  a  lighter  colour  than  the 
main  mass  of  the  sediment  ". 

The  only  previous  work  on  trace-fossils  from  the  Lower  Chalk  is  that  of  Davies 
(1879)  and  Bather  (1911).  The  terebellids  described  by  these  authors  are,  in  part, 
burrows,  whilst  the  latter  described  a  single  fragment  referred  to  as  Keckia  (?)  sp. 

III.     LOCATION  OF  SPECIMENS 

The  author's  collection  and  the  types  of  "  Terebella  "  cancellata  Bather  and  "  T." 
harefieldensis  White  are  in  the  collections  of  the  British  Museum  (Natural  History), 
Dr.  R.  P.  S.  Jefferies'  collection  is  in  the  Sedgwick  Museum,  Cambridge.  These  are 
abbreviated  to  B.M.  (N.H.),  and  S.M.C.  respectively  in  the  following  account. 

IV.     SYSTEMATIC  DESCRIPTIONS 

Ichnogenus  THALASSINOIDES  Ehrenberg  1944 

Type  species.  By  the  original  designation  of  Ehrenberg  (1944)  Thalassinoides 
callianassae  Ehrenberg,  from  the  Miocene  (Burdigalian)  of  Burgschleinitz,  Eggenberg, 
Austria. 

Discussion.  This  trace  fossil  genus  was  erected  for  a  ramifying  system  of  cylindri- 
cal burrows  from  Miocene  sands,  intimately  associated  with,  and  probably  formed  by 
crustaceans  (identified  as  Callianassa  sp.)  described  by  Ehrenberg  six  years  previously 
(1938).  The  original  diagnosis  is  as  follows:  "  Die  Gattung  ,, Thalassinoides  "  ware 
wie  folgt  zu  kennzeichnen :  Gange  und  Gangsysteme  oder  bzw.  deren  Ausfiillungen 
(Kerne)  mit  mehr  oder  weniger  Y-formigen  Gabelungen  oder  Verzweigungen,  meist 
ohne  wesentliche  Oberflachenskulpturen :  sonstige  Form  und  Durchmesser  merklich 
wechselnd. 

"Typus-,,  Art  "  Th.  callianassae  mit  den  Charakteren  der  ,,  Gattung  "  aus  dem 
Burdigal  von  Burgschleinitz  bei  Eggenberg.     N  :  D.  Typusexemplar  das  im  Palaon- 


i32  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

tolog.  u.  Palaobiolog.  Institut  der  Universitat  Wien  verwahrte  Urstuck  zu  Ehrenberg 
1938  Tafel  28,5." 

A  brief  diagnosis  in  English  has  been  given  by  Hantzschel  (1962).  On  the  basis 
of  the  present  material  it  may  be  emended  as  follows: 

Extensive  burrow  systems  with  both  vertical  and  horizontal  elements.  Burrows 
cylindrical,  between  2  and  20  cm.  in  diameter.  Branching  regular,  characterized  by 
Y-shaped  bifurcations,  swollen  at  point  of  branching.  Horizontal  elements  joining 
to  form  polygons.  Burrow  dimensions  variable  within  a  system.  Horizontal 
systems  connecting  to  surface  by  vertical  or  steeply  inclined  shafts,  widely  associated 
with  callianassid  remains. 

Hantzschel  (1962,  1965)  regards  Vomacispongites  de  Laubenfels  (1955  :  108)  as  a 
synonym  of  Thalassinoides .  Vomacispongites  was  introduced  by  de  Laubenfels  as  an 
"  unrecognizable  supposed  sponge  ",  as  follows: 

"  Vomacispongites  de  Laub.  nom.  nov.  (pro  Spongites  Schloth.  1820  (non.  Oken 
1814))  ",  the  type  species  is  Spongites  pertusus  Schlotheim  (1820  :  369)  based  on  a 
specimen  from  a  Cretaceous  chert  from  Amberg  (W.  Germany),  compared  by  von 
Schlotheim  to  Spongia  pertusa  Esper  (Esper  1799  :  246-7,  pi.  26,  figs.  1,  2).  Esper's 
figure  is  clearly  a  sponge,  and  I  can  only  presume  that  Hantzschel  has  examined  the 
original  specimen,  since  the  original  description  does  not  suggest  a  Thalassinoides. 
The  genus  Aschemonia  Dettmer  (1914)  is  too  poorly  defined  for  comparison,  but  may 
well  be  a  Thalassinoides. 

In  addition  to  the  association  of  Thalassinoides  with  Callianassa  sp.  recorded  by 
Ehrenberg  (1938),  Glaessner  (1947)  describes  what  are  clearly  Thalassinoides  in 
association  with  callianassids  from  the  Eocene  of  Victoria  (Australia),  whilst  Mertin 
(1941)  records  Protocallianassa  in  association  with  what  are  probably  Thalassinoides 
in  the  Upper  Cretaceous  of  Germany.  Hantzschel  (1965),  Seilacher  (1955,  1964), 
Farrow  (1966)  and  Hallam  (1961)  all  regard  Thalassinoides  as  a  crustacean  burrow. 

Thalassinoides  is  very  widespread,  and  has  been  recorded  from  the  Trias  (Reis 
1910,  Fiege  1944),  Lias  (Rieth  1932,  Seilacher  1955,  Hallam  1961),  Oxfordian  (Wilson 
1949),  Portlandian  (Pruvost  &  Pringle  1924,  Arkell  1935),  Cretaceous  (Geinitz  1842 
etc.)  and  Tertiary  (Ehrenberg  1938,  Glaessner  1947).  The  geographical  range  of 
this  form  covers  Europe,  Asia  and  Australia. 

In  Britain,  this  trace  fossil  has  been  recorded  from  the  Lower  Lias  by  Hallam  (1961), 
whilst  the  fucoids  recorded  by  Blake  &  Hudleston  (1877  :  271)  and  Arkell  (1936  :  63) 
from  the  Oxfordian  (Corallian,  Nothe  Grits)  of  the  Dorset  coast,  andfigured  from  a  similar 
horizon  in  Yorkshire  by  Wilson  (1949  :  256,  pi.  10)  are  clearly  Thalassinoides,  as 
are  the  fucoids  figured  by  Arkell  (1925  :  pi.  22,  a)  from  the  Portlandian  (Portland 
Sand,  Black  Sandstones)  of  the  Dorset  coast.  Farrow  (1966)  records  it  from  many 
levels  in  the  Yorkshire  Jurassic. 

I  have  noted  this  trace  fossil  at  many  horizons  and  localities:  Triassic:  Rhaetic, 
South-Devon  coast  between  Seaton  and  Lyme  Regis,  piping  the  basal  bone  bed  into 
the  underlying  Keuper  (Text-fig.  2,  G.).  Jurassic:  The  whole  of  the  Dorset  Lias 
(Text-figs.  1,  g-j;  2,  f).  Cretaceous:  Lower  Greensand,  Folkestone  beds  at  Folke- 
stone (Text-fig.  2,  i-k)  associated  with  Gyrolithes  type  structures;  Upper  Greensand 
of  Southern  England  and  throughout  the  whole  of  the  Lower  Chalk  and  in  the  Middle 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 


133 


Fig.  1.  a.  Thalassinoides  sp.  Lower  Chalk,  PUpper  Cenomanian;  White  Nothe, 
Dorset.  Plan,  x  ^.  b,  c.  Thalassinoides  saxonicus  (Geinitz).  Lower  Chalk,  Middle 
Cenomanian,  Chalk  below  Totternhoe  Stone;  Houghton  Regis,  near  Dunstable,  Beds. 
Plan,  x  tV-  d,  f.  Laminated  structures.  Lower  Chalk,  Middle  Cenomanian ;  near 
Beachy  Head,  Eastbourne,  Sussex,  Plan,  x  -fa.  e.  Laminated  structure.  Lower 
Chalk,  Middle  Cenomanian,  bed  7;  Folkestone,  Kent.  Plan,  X -3^-.  g— J.  Thalas- 
sinoides sp.  Upper  Lias,  Toarcian;  near  Seatown,  Dorset.  Plan,  X -^-.  k.  Thalas- 
sinoides sp.     Upper  Greensand;    Foxmould,  Humble  Point,  South  Devon,     Plan,   X  ^-, 


134  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

and  Upper  Chalk  where  burrows  are  preserved  (i.e.  the  Melbourn  Rock  (Lower 
Turonian)  and  below  the  Chalk  Rock  (Upper  Turonian)).  At  least  four  forms  can  be 
recognized  in  the  Lower  Chalk. 

Thalassinoides  saxonicus  (Geinitz) 
(PI.  i,  fig.  i;  PI.  5,  figs.  2,  3;  PI.  6,  figs.  3,  4;  Text-figs,  1,  b,  c;  2,  e) 

1842     Spongites    saxonicus    Geinitz  :  96,    pi.    12,    fig.    1    only  (fig.    2  =  Ophiomorpha  nodosa 
Lundgren) . 

1852  Spongites  saxonicus  Geinitz;   von  Otto  :  20,  pi.  6,  figs.  2,  3. 

1871  Spongites  saxonicus  Geinitz;    Geinitz  :  21,  pi.  1,  figs.  1-5  only. 

1878  Spongia  saxonica  Geinitz;    Fric  :  149. 

1878  Spongites  gigas  Fric:   75,  149. 

1885  Spongites  saxonicus  Geinitz;    Pocta  :  30. 

1899  Spongites  saxonicus  Geinitz;    Semenow  :  6. 

1909  Cylindrites  spongioides  Goeppert  emend.  Richter  :  8,  11. 

1912  Spongites  saxonicus  Geinitz;    Dettmer  :  1 14-126  (pars.),  ?pl.  8,  figs.  4-6. 

?igi4  Aschemonia  gigantea  Dettmer  :  287,  fig. 

1915  Spongites  saxonicus  Geinitz;    Dettmer  :  285-287  (pars.). 

?ig28  Spongites  sp.  Lamprecht  :  8,  9,  pi.  2. 

1932  Spongites  saxonicus  Geinitz;    Rieth  :  30,  pi.  5a,  1,  2. 

1934  Spongites  saxonicus  Geinitz;    Andert  :  68. 

1934  Spongites  saxonicus  Geinitz;    Hantzschel  :  313. 

1944  Spongites  saxonicus  Geinitz;    Fiege  :  419. 

1952  Spongites  saxonicus  Geinitz;    Hantzschel  :  146. 

1954  Cylindrites  saxonicus  Prescher  :  59,  text-fig.  19. 

?i955  Spongites  sp.,  Seilacher  :  text-fig.  5,  98. 

1962  Spongites  saxonicus  Geinitz;    Hantzschel  :  218. 

1965  Spongites  saxonicus  Geinitz;   Hantzschel  :  88. 

1967  "  Spongites  "  saxonicus  Geinitz;    Kennedy  :  368 

Diagnosis.  Thalassinoides  with  horizontal  tunnels  between  5  and  20  cm.  in 
horizontal  diameter.  System  very  extensive,  tunnels  joining  to  form  huge  polygons 
up  to  60  cm.  across,  connected  to  surface  by  short  vertical  shafts.  Surface  of  burrow 
mamillated,  individual  mounds  5  to  10  mm.  long,  elongated  parallel  to  length  of 
tunnel. 

Lectotype.  Here  designated,  the  original  of  Spongites  saxonicus  Geinitz  1842, 
pi.  22,  fig.  1  only:   Upper  Cretaceous;   Germany. 

Description.  Systems  arise  from  short,  vertical  shafts,  equal  in  diameter  to  the 
widest  part  of  the  horizontal  elements,  which  are  up  to  40  cm.  below  the  surface. 
The  horizontal  tunnels  are  elliptical  in  section  and  at  a  single  level.     Tunnel  diameters 

Fig.  2.  a-d.  Thalassinoides  visurgiae  Fiege  (after  Fiege  1944)  X  \.  e.  Spongites 
saxonicus  Geinitz,  Sketch  of  lectotype  (after  Geinitz  1842)  xj.  r.  Thalassinoides  sp. 
Lower  Lias;  Pinhay  Bay,  S.  Devon.  Plan,  x  -fe.  g.  Thalassinoides  sp.  Top  of 
Keuper,  full  of  Rhaetic  Bone  Bed;  Charlton  Bay,  S.  Devon.  Plan,  x  \.  H.  Thalas- 
sinoides sp.  Upper  Chalk,  Upper  Turonian,  Chalk  Rock;  Hitch  Wood,  near  Hitchin, 
Herts.  Plan  view  of  3-dimensional  tunnel  system  beneath  the  Chalk  Rock  hardground. 
Xj.  i-k.  Thalassinoides  sp.  Lower  Greensand,  Folkestone  Beds,  Lower  Albian; 
Copt  Point,  Folkestone,  Kent.     Plan,   x  -fe. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 


135 


J 


136  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

vary  between  20  by  13  cm.  in  the  centre  of  systems,  to  5  by  3-5  cm.  at  the  extremities. 
Branching  is  very  regular,  at  intervals  of  about  30  cm.  tunnels  bifurcate,  with  a 
symmetrical  Y-shaped  junction,  slight  widening  giving  rounded  angles.  The  overall 
pattern  is  of  large  polygons  up  to  60  cm.  across  (Plate  1,  fig.  1).  Specimens  showing 
terminations  are  uncommon;  the  ending  figured  (Plate  5,  fig.  2)  is  swollen,  measuring 
12  by  5  by  3-5  cm.  at  the  end  of  a  3  by  2-5  cm.  tunnel.  The  outer  surfaces  of  tunnels 
are  covered  by  low  mounds:  on  small  specimens  these  vary  from  17-5  to  18  mm.  long 
by  7-5  to  11*5  mm.  wide  by  3-5  to  6-5  mm.  high:  on  larger  specimens  16-18  mm.  by 
9-12  mm.  by  4-6  mm.  The  arrangement  of  these  mounds  shows  little  regularity 
other  than  a  preferred  orientation  parallel  to  the  length  of  the  tunnel.  These 
structures  are  rarely  preserved  on  bottom  surfaces,  which  are  covered  by  a  felted  mass 
of  Chondrites  burrows  (PI.  5,  fig.  3). 

Sections  of  tunnels  suggest  that  some  of  these  mounds  are  discrete,  with  a  definite 
separation  from  the  burrow  filling.  The  majority  have  only  a  suggestion  of  a  plane 
of  separation.  In  addition  to  mounds,  oval  depressions  of  a  similar  size,  surrounded 
by  a  raised  rim  are  present,  as  are  ridges,  generally  20  mm.  long  and  2  mm.  high, 
running  between  the  mounds. 

Discussion.  The  lectotype,  as  figured  by  Geinitz,  is  a  large,  branching  cylindrical 
body  with  a  maximum  diameter  of  5  cm.  The  surface  is  covered  by  small  mounds, 
5  mm.  long  and  2  to  3  mm.  wide,  elongated  parallel  to  the  length  of  the  branches. 
A  smooth  half  cylinder  5  mm.  in  diameter  runs  along  the  centre  of  the  main  part, 
joining  with  similar  bodies  on  the  branches  (Text-fig.  2,  e).  The  other  specimen 
figured  by  Geinitz  (pi.  22,  fig.  2)  can  clearly  be  referred  to  Ophiomorpha  nodosa 
Lundgren. 

Geinitz  regarded  Spongites  saxonicus  as  a  horny  sponge  (Ceratospongidae) ,  a  view 
also  held  by  von  Otto  (1854),  Fric  (1878),  Pocta  (1885)  and  many  other  early  workers. 
Goeppert  (1842  :  115,  pi.  46,  figs.  1-5,  pi.  48,  figs.  1,  2)  described  what  he  regarded  as 
a  fossil  alga,  Cylindrites  spongioides,  and  he  subsequently  (1847)  considered  this 
species  to  have  priority  over  S.  saxonicus,  and  that  both  were  algae.  Cylindrites 
spongioides,  as  originally  proposed,  includes  a  number  of  different  trace-fossils. 
Forms  figured  on  plate  46,  figs.  1-4  are  simple  crustacean  burrows  (type  B,  p.  47), 
or  possibly  Ophiomorpha.  The  other  specimens  (pi.  46,  fig.  4;  pi.  48,  figs.  1,  2)  are 
smooth  cylindrical  burrows  with  swollen  portions,  differing  from  S.  saxonicus  in 
smaller  size  and  lack  of  ornament. 

Cylindrites  has  been  used  by  many  authors  for  fucoids  (Eichwald  1865,  Watelet 
1866  etc.)  or  trace  fossils  (Prescher  1954),  but  is  not  available  due  to  prior  usage  by 
Gmelin  (1793)  and  Sowerby  (1825)  as  gastropod  genera.  Richter's  (1909)  emendation 
of  Cylindrites  is  unfortunate,  for  he  clearly  includes  large  burrows  (up  to  15  cm.  in 
diameter),  probably  Thalassinoides  saxonicus,  specimens  of  Ophiomorpha  nodosa 
(pi.  9,  fig.  7,  pi.  12,  fig.  5,  pi.  13,  fig.  6),  plant  debris  (pi.  12,  figs.  1,  2,  pi.  13,  fig.  6) 
and  smooth  burrows  with  swollen  portions  (pi.  9,  figs.  1-2). 

As  pointed  out  by  Hantzschel  (1952),  Cylindrites  spongioides  may  be  a  synonym  of 
Halymenites  cylindricus  Sternberg: 

"  H,   fronde  fistulosa   terente   pinnatim   ramosa,   ramus   opposites   simplicibus 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  137 

patenibus  cylindricus  obtusis,  terminale  longiors  ...  In  schisto  saxi  arenaci  Pirnensis 
(Greensand  anglorum)  prope  Tetschen  ad  albim  Bohemae."  (Sternberg  1833.) 

Until  the  branching  pattern  of  C.  spongioides  is  described,  it  is  not  possible  to 
decide  if  it  is  a  form  of  Thalassinoides,  although  it  is  clear  that  it  is  a  smooth  crusta- 
cean burrow,  the  swollen  portions  representing  "  turn-arounds  ",  comparable  to  the 
burrows  produced  by  the  living  crustaceans  Upobegia  pugettensis  (Dana)  (MacGinitie 
1930)  and  Callianassa  califomiensis  Dana  (MacGinitie  1934,  Stevens  1928).  These 
burrows  were  noted  as  early  as  1760  by  Schulze  (41-46,  pi.  2,  figs.  1-5)  who  regarded 
them  as  crinoid  remains.  In  view  of  their  interpretation  as  callianassid  burrows  it  is 
interesting  to  note  that  Goeppert  (1854)  recorded  C.  spongioides  in  association  with 
remains  of  Protocallianassa  antiquua  (von  Otto) . 

Dettmer  (1912)  regarded  Spongites  saxonicus  as  a  giant  foraminiferan! 

Hantzschel  (1934,  1935)  records  Xenohelix  saxonica  Hantzschel  {^Gyrolithes 
Saporta)  associated  with  Spongites  saxonicus;  spiral  structures,  perhaps  Gyrolithes 
occur  in  the  Folkestone  beds  (L.  Albian)  at  Folkestone,  Kent  in  association  with 
Thalassinoides,  and  have  been  recorded  associated  with  Ophiomorpha  in  the  Miocene 
of  Borneo  (Keij  1965)  and  elsewhere  (Kilpper  1962).  This  type  of  association  indicates 
the  artificial  nature  of  trace-fossil  taxa,  as  it  suggests  that  Gyrolithes,  Thalassinoides 
and  Ophiomorpha  are  all  synonymous,  the  first  having  priority. 

The  best  preserved  examples  of  T.  saxonicus  I  have  found  are  from  beneath  the 
Totternhoe  Stone  (Middle  Cenomanian)  of  the  Chilterns,  particularly  Houghton 
Regis  (Bedfordshire).  Here,  hard,  gritty  Totternhoe  Stone  is  piped  into  the  very 
soft  chalk  below  and  the  burrows  so  filled  can  be  completely  freed  of  matrix  (PI.  5, 
figs.  2,  3;   PL  6,  figs.  3,  4). 

The  presence  of  phosphatic  pebbles  and  shells  in  these  burrows  indicates  that  they 
were  open  on  the  sea  floor,  and  were  filled  passively,  probably  after  being  vacated. 
Individual  systems  extend  over  several  square  metres  and  indicate  firm  sediment,  as 
I  have  never  seen  signs  of  collapse  into  them. 

The  ridges  on  the  outer  surface  are  interpreted  as  scratches  produced  by  the 
inhabitant  whilst  digging  or  moving  through  the  system ;  the  mamillated  surface  as 
a  result  of  worked  pellets  pushed  into  the  wall  of  the  tunnel  and  smoothed  off  or 
worn  smooth  by  the  passage  of  the  animal's  body.  Pellets  are  impressed  into  burrow 
walls  in  this  manner  by  the  crustacean  Callianassa  major  Say  (MacGinitie  in  Hantz- 
schel 1952,  Weimer  &  Hoyt  1964).  The  oval  depressions  with  their  surrounding 
ridges  appear  to  be  the  sites  of  pellets  of  soft  chalk  which  have  been  washed  away  in 
preparation.  The  most  likely  purpose  of  these  pellets  is  to  support  the  burrow  walls, 
a  procedure  used  by  living  callianassids  (Pohl  1946) ;  swollen  portions  at  points  of 
branching  and  burrow  terminations  are  comparable  with  the  "  turn  arounds  "  of 
burrows  of  this  group  (MacGinitie  1930,  1934,  Pohl  1946).  All  the  features  of 
T.  saxonicus  are  thus  comparable  with  Recent  callianassid  burrows.  This  view  is 
enhanced  by  the  presence,  in  the  infilling  of  T.  saxonicus,  of  rod-like  phosphatized 
faecal  pellets  (type  A  of  Wilcox  1953),  more  abundant  than  elsewhere  in  the  Lower 
Chalk,  which,  from  the  presence  of  internal  canals,  are  diagnostic  of  anomurans 
(Wilcox  1953,  Moore  1932). 

Internally,  these  burrows  show  intense  re-working  (see  p.  149).   A  puzzling  feature 


138  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

of  previously  described  specimens  of  T.  saxonicus  is  the  small  burrow  running  down 
the  centre  of  the  lower  surface  of  the  system.  In  the  lectotype  this  is  very  regular, 
but  in  other  specimens  it  clearly  strays  from  the  mid-line.  Geinitz  (1842)  interpreted 
this  as  a  juvenile  sponge.  My  own  material  suggests  that  this  is  another,  smaller 
species  of  Thalassinoides,  which  sometimes  follows  the  mid-line  of  the  bottom  of 
the  larger  burrow,  but  which  often  leaves,  passing  out  into  the  surrounding  sediment 
(PL  6,  fig.  4).  These  smaller  burrows  may  be  the  products  of  the  juveniles  of  the 
T.  saxonicus  animal,  but  as  I  have  never  seen  transitions  it  is  regarded  as  a  distinct 
form,  T.  ornatus  nov.  (p.  141). 

Whilst  most  systems  correspond  to  the  above  description,  occasionally  tunnels 
are  found  filled  with  coarse,  sandy  chalk  made  up  of  shell  fragments  and  microfossils. 
This  material  represents  the  remains  of  the  burrow  filling  after  the  inhabitant  has 


B  C 

Fig.  3.  Alternative  interpretations  of  laminated  structures,  a.  As  surface  trace;  b. 
As  partially  filled  burrow;  c.  As  totally  filled  burrow  with  semi-circular  section. 
All  xj. 

sifted  out  the  finer  portion  for  ingestion;  the  faecal  pellets  associated  with  these 
burrows  show  only  fine-grained  material  when  sectioned  (Wilcox  1953),  suggesting 
this  mode  of  feeding.  Presumably  the  coarser  debris  was  normally  removed  from 
the  system  and  washed  away  by  bottom  currents,  being  only  occasionally  packed  into 
a  disused  part  of  the  burrow.  In  a  single  instance,  from  the  Upper  Cenomanian  at 
Dover,  a  large  mass  of  this  coarse  debris,  lying  above  a  T.  saxonicus  system  seemed 
to  represent  material  dumped  outside  the  burrow  opening  adjoining  the  entrance 
shaft. 

Thalassinoides  saxonicus  and  "  Laminated  structures  ".  I  have  used  the  name 
"  laminated  structures "  (Kennedy  1967)  for  problematic  structures  occurring 
throughout  the  Lower  Chalk  which,  in  section,  show  fine,  horizontal  laminations 
made  prominent  by  their  resistance  to  weathering  (due  to  a  calcite  cement)  and  brown 
colour  (due  to  disseminated  limonite) .  In  plan,  these  structures  show  a  form  identical 
in  size-range  and  mode  of  branching  with  T.  saxonicus  (Text-fig.  1,  d-f).  In  section 
they  can  be  described  in  terms  of  a  continuous  series  defined  by  two  end  members: 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  139 

(i)  Horizontal  bands,  up  to  60  cm.  long  and  5-10  cm.  deep,  laminated  throughout, 
the  laminations  essentially  parallel. 

(ii)  Semi-circular  sections,  the  curved  surface  downwards,  with  diameters  from 
10  cm.  upwards.  With  increasing  diameter  the  radius  of  curvature  of  the  lower 
surface  increases,  passing  into  form  (i).  Laminations  are  essentially  parallel, 
with  a  tendency  to  dip  towards  the  centre  of  the  section. 

In  both  cases,  weathered  and  cut  sections  show  that  the  laminations  consist  of 
alternations  of  normal  chalk  and  coarse  debris  of  shell  and  echinoderm  fragments, 
foraminifera  and  other  sand  grade  microfossils,  cemented  by  calcite  and  coloured 
brown  by  small  quantities  of  limonite,  perhaps  after  pyrite.  The  laminations  are 
between  2  and  20  mm.  thick,  and  may  vary  laterally  (PI.  8,  fig.  3).  In  thin  section, 
these  structures  show  very  ill-defined  graded  bedding,  the  coarse  material  grading  up 
into  the  normal  chalk  above. 

These  laminated  structures  are  cut  by  burrows  (PI.  8,  fig.  3;  PI.  5,  fig.  1)  indicating 
a  primary  origin.  Whilst  the  general  sense  of  the  laminations  is  horizontal  and  paral- 
lel, corrugations  and  other  disturbances  are  common  (PI.  8,  fig.  3).  Some  of  these 
structures  are  clearly  the  result  of  subsequent  burrowing  (PI.  8,  fig.  3),  whilst  other 
irregularities  appear  to  be  the  result  of  slumping  of  the  layers  (PI.  8,  fig.  3).  Bottom 
surfaces  are  rather  irregular  (PI.  8,  fig.  3),  in  part  as  a  result  of  burrows  along  the  basal 
interface. 

Interpretation.     In  view  of  the  similarity  in  size-range  of  these  structures  and 
Thalassinoides  saxonicus  and  the  identical  branching  pattern,  they  are  clearly  the 
result  of  the  activities  of  the  same  organisms;  crustaceans.     Whereas  T.  saxonicus 
is  clearly  a  burrow,  elliptical  in  section,  laminated  structures  generally  have  a  flat 
top.     Three  interpretations  are  possible  (Text-fig.  3). 
(i)  They  are  the  filling  of  the  lower  parts  of  burrows, 
(ii)  They  are  completely  filled  burrows  semi-circular  in  section, 
(iii)  They  are  a  surface  trace. 

I  have  examined  many  examples  in  the  field;  most  show  no  indications  of  an 
associated  burrow.  A  few  show  what  could  be  interpreted  as  the  upper  part  of  a 
burrow,  but  at  present  the  evidence  suggests  they  were  a  surface  trace,  although  the 
relationship  seen  in  Plate  8  could  be  interpreted  as  the  intersection  of  two  burrows 
with  a  semicircular  section,  completely  full  of  laminated  sediment. 

The  laminations  are  interpreted  as  the  result  of  sifting  of  the  sediment  by  the  ani- 
mals producing  these  structures.  As  already  indicated  (p.  137)  faecal  material  suggests 
they  lived  on  the  finer  fractions;  the  coarse  layers  are  the  remains  left  after  this 
sifting.  Whilst  this  can  explain  the  formation  of  one  layer,  I  can  offer  no  explanation 
of  the  repeated  alternation  of  coarse  and  fine  layers. 

Explanation  of  these  structures  as  a  feeding  trace  of  Teichichnus  type  (Seilacher 
1955,  Hantzschel  1962)  is  unsatisfactory  due  to  the  absence  of  an  obvious  burrow  in 
association,  unless  the  initial  burrow  were  very  shallow  and  invariably  broke  the 
sediment-water  interface.  An  inorganic  origin — that  these  are  Thalassinoides, 
exposed  by  erosion  and  filled  by  swept-in  coarse  material  alternating  with  fine  mud 
deposited  by  gravitational  settling  or  other  currents — is  rejected;  other  hollows  on  the 


i4o  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

sea  floor  lack  a  laminated  fill,  whilst  one  would  expect  to  find  truncated  Thalassinoides 
without  a  laminated  fill,  which  I  have  never  seen. 

Occurrence.  Solid  specimens  of  T.  saxonicus  occur  abundantly  beneath  the 
Totternhoe  Stone  in  the  Chilterns.  Large  Thalassinoides,  identical  in  size  and  mode 
of  branching  are  common  in  all  coastal  sections  whilst  large  oval  burrow  sections  are 
abundant  in  all  sections  and  are  regarded  as  identical  with  T.  saxonicus.  A  large 
species  of  Thalassinoides  occurs  beneath  the  Chalk  Rock  (Text-fig.  2,  h)  but  differs 
from  T.  saxonicus  in  having  three-dimensional  tunnels.  These  were  described  by 
Billinghurst  (1927)  as  "  solution  channels  ". 

T.  saxonicus  is  widely  recorded  from  the  Cretaceous  of  Germany  and  Central 
Europe.  The  specimen  of  Thalassinoides  figured  by  Seilacher  (1955)  from  the 
Tertiary  may  belong  to  this  form.  "  Laminated  structures  "  are  common  through- 
out the  whole  of  the  Lower  Chalk,  particularly  in  the  Middle  Cenomanian.  At 
Folkestone  (Kent)  bed  7  (Jukes-Browne  &  Hill  1903)  can  be  traced  all  along  the 
coast,  even  when  high  in  the  cliffs,  because  of  the  abundance  of  these  structures. 


Thalassinoides  cf.  suevicus  (Rieth) 
(PI.  1,  fig.  2) 

1932  Spongites  suevicus  quenstedti  Rieth  :  274. 

1932  Spongites  suevicus  Quenstedt;    Rieth  :  292. 

1932  Cylindrites  suevicus  (Quenstedt)  Rieth,  pi.  13a,  b. 

?i944  Thalassinoides  visurgiae  Fiege  :  416-421,  424,  text-fig.  4. 

1955  Spongites  suevicus  Quenstedt;    Seilacher,  text-figs.  5,  57. 

?ig64  Thalassinoides  sp.  Hantzschel  :  302,  pi.  14,  fig.  3. 

1964  Thalassinoides  suevicus  (Rieth);    Hantzschel  :  302. 

A  Thalassinoides  with  tunnel  diameters  between  2  and  5  cm.  is  occasionally  seen  in 
alien  blocks  of  Upper  Greensand  and  Lower  Chalk  at  many  coastal  sections  (East- 
bourne, Compton  Bay,  etc.).  Those  in  the  Upper  Greensand  in  part  arise  from  the 
base  of  the  Glauconitic  Marl. 

In  size,  mode  and  angle  of  branching  these  are  comparable  with  "  Spongites  " 
suevicus  Rieth,  from  the  Lias  and  Dogger  of  Germany  (as  pointed  out  by  Hantzschel 
(1964  :  302)  this  name  must  be  attributed  to  Rieth).  There  is  also  a  strong  resem- 
blance to  the  fragment  figured  by  Hantzschel  (1964)  from  the  Campanian  of  Beckum 
(Westphalia) . 

Thalassinoides  visurgiae  Fiege,  from  the  Trias  (Muschelkalk)  of  North  Germany, 
is  based  on  branching  portions  (Text-fig.  2,  a-d)  and  appears  identical  with  T. 
suevicus. 

T.  cf.  suevicus  differs  from  T.  saxonicus  in  its  much  smaller  size  and  absence  of 
ornamentation.  It  is  not  referred  definitely  to  T.  suevicus  because  of  the  poor 
preservation.  The  systems  are  horizontal  as  far  as  has  been  seen.  It  is  not  clear 
how  much  of  the  piping  beneath  the  Glauconitic  Marl  is  due  to  this  form  (PL  1,  fig.  3), 
but  attitude  and  tunnel  diameters  are  comparable. 

Specimens  of  Thalassinoides  from  the  Upper  Greensand  and  from  beneath  the 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  141 

Totternhoe  Stone  at  Hunstanton,  agreeing  in  size  with  this  form  occasionally  show 
internal  structure,  seen  in  section  as  concavo-convex  laminations  (Text-fig.  6,  d). 
This  type  of  burrow  filling  is  discussed  on  page  161. 

Occurrence.  Fairly  frequent  in  the  whole  of  the  Lower  Chalk  of  southern  Eng- 
land. Comparable  forms  occur  in  the  Lias  and  Upper  Greensand :  Lias  and  Dogger 
of  Germany  and  Campanian  of  Bochum,  Germany. 


Thalassinoides  ornatus  ichnosp.  nov. 
(PL  6,  fig.  4;   PL  7,  fig.  6) 

Diagnosis.  Small  Thalassinoides,  tunnel  diameters  between  16  by  8  mm.  and 
22  by  10  mm.  System  largely  horizontal,  surface  of  tunnels  covered  in  reticulate 
ridges. 

Holotype.     B.M.  (N.H.)  T.559.     Paratype  B.M.  (N.H.)  T.551. 

Locality  and  horizon.  The  holotype  (associated  with  T.  saxonicus)  is  from  the 
Lower  Chalk  immediately  beneath  the  Totternhoe  Stone  at  Houghton  Regis  (Bedford 
shire)  (National  Grid  Reference  T.L. 013233)  and  is  Middle  Cenomanian  in  age.  The 
paratype  is  from  the  same  horizon  and  locality.  This  species  is  not  uncommon 
beneath  the  Totternhoe  Stone  elsewhere  in  the  Chilterns :  poorly  preserved  material 
from  the  Lower  Chalk  of  the  Weald  may  also  belong  to  this  form. 

Description.  The  tunnels  are  generally  horizontal  or  gently  inclined,  with 
typical  Thalassinoides  branching  pattern.  Individual  tunnels  are  oval  in  section, 
dimensions  varying  between  16  by  8  mm.  to  22  by  10  mm.  The  whole  surface  is 
covered  in  delicate  intersecting  ridges  (PL  7,  fig.  6) ;  some  tunnels  are  gently  curved. 
Branching  points  are  swollen,  whilst  swollen  portions  with  diameters  of  about  three 
times  that  of  the  adjoining  tunnel  are  present. 

Discussion.  This  form  is  quite  common  beneath  the  Totternhoe  Stone  in  Bed- 
fordshire, often  following  the  mid-line  of  the  lower  surface  of  T.  saxonicus  burrows 
(PL  6,  fig.  3).  As  already  suggested  (p.  138)  this  may  be  the  explanation  of  the  cylin- 
drical central  body  figured  by  previous  workers  (Geinitz  1842,  1871,  von  Otto  1854, 
Seilacher  1955  etc.).  This  pattern  is  not  regular,  the  smaller  burrows  often  passing 
through  the  larger  burrows  and  occurring  in  the  surrounding  sediment. 

Interpretation  of  T.  ornatus  as  the  work  of  juveniles  of  T.  saxonicus  is  considered 
unlikely  in  the  absence  of  intermediate  forms. 

As  with  the  forms  of  Thalassinoides  already  discussed,  the  features  of  T.  ornatus 
agree  with  an  interpretation  as  crustacean  burrows:  reticulate  surface  ridges  are 
scratches  on  the  inside  of  the  burrow  produced  during  digging  or  when  moving  through 
the  system,  the  swollen  portions  are  clearly  "  turn-arounds  ". 

The  surface  ornament  of  T.  ornatus  resembles  that  on  Spongeliomorpha  (p.  151), 
also  regarded  as  a  crustacean  burrow.  The  two  forms  are  distinguished  by  the  more 
regular  ornamentation  and  branching  of  T.  ornatus.  This  form  differs  from  T.  cf . 
suevicus  by  the  presence  of  a  reticulate  ornamentation  and  swollen  "  turn- 
arounds ".     Clearly,  with  poorly  preserved  material  the  two  forms  may  be  confused. 

geol.  15,3.  15 


i42  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Thalassinoides  paradoxica  (Woodward) 
(PL  3;   PL  4;   PL  8,  fig.  5;   PL  9,  %•  2;  Text-figs.  4,  5,  a-b) 

1814  .   .  .  singular  organic  body  .   .   .,    Webster,  pi.  27,  fig.  1. 

1823  .   .   .a  remarkable  ramifying  zoophyte  .   .   .,    Taylor  :  82. 

1830  Spongia  paradoxica  Woodward  :  5. 

1833  Spongia  paradoxica  Woodward;   Woodward  :  29,  30,  54. 

1835  ...  a  ramose  zoophyte  .   .  .  Rose  :  54,  275,  276. 

1859  Spongia  paradoxica  Woodward;   Wiltshire  :  275,  277,  pi.  1,  figs.  1,  2. 

1864  Spongia  paradoxica  Woodward;    Seeley  :  331. 

1869  Siphonia  paradoxica  (Woodward)  Wiltshire  :  176. 

1871  Problematicum,  Geinitz,  pi.  38,  fig.  8. 

1884  Spongia  paradoxica  Woodward;   Hughes  :  273-279. 

1899  Spongia  paradoxica  Woodward;   Whitaker  &  Jukes-Browne  :  36,  55. 

1900  Spongia  paradoxica  Woodward;    Jukes-Browne  &  Hill  :  303. 
1903  .   .   .  stems  of  Siphonia  .   .   .  ,  Jukes-Browne  &  Hill  :  209. 
1932  Problematicum,  Rieth,  text-fig.  35  (after  Geinitz). 

1961  "  Spongia  paradoxica  "  Woodward;   Peake  &  Hancock  :  301,  330. 
?ig6i     "  Spongia  paradoxica  "  Woodward;    Rios  &  Hancock,  pi.  16. 

1962  Spongia  paradoxica  Woodward;    Hantzschel  :  W  242. 

Diagnosis.  Medium  sized  Thalassinoides,  with  irregular,  very  extensive  horizon- 
tal burrow  network,  occurring  at  several  levels,  connected  by  vertical  shafts. 
Diameter  of  tunnels  variable,  between  7  and  60  mm.,  short  blind  tunnels  very  com- 
mon. Surface  covered  with  longitudinal  ridges.  Generally  occurs  associated  with 
erosion  surfaces. 

Neotype.  Here  designated,  B.M.  (N.H.)  T.545  from  the  Paradoxica  bed,  base  of 
Lower  Chalk  (Lower  Cenomanian) ;  Hunstanton  Cliff,  Hunstanton,  Norfolk. 

Description.  This  is  the  most  irregularly  branching  Thalassinoides  I  have  seen. 
The  burrows  have  an  irregular  section,  and  may  be  depressed  or  rounded-angular, 
varying  in  a  single  system  between  7  and  60  mm.  in  diameter.  A  large  tunnel  may 
give  rise  to  a  side  branch  less  than  a  quarter  of  its  own  diameter.  As  in  other  species 
of  Thalassinoides,  the  principal  element  of  branching  is  a  Y  fork,  with  an  increase 
in  diameter  around  the  point  of  branching,  the  tunnel  tending  to  widen  between  the 
forks  of  the  Y.  Distance  between  branching  points  is  very  variable,  between  1  and 
20  cm.  Many  of  the  branches  terminate  after  short  distances,  giving  the  system  an 
antler-like  appearance,  whilst  at  every  point  on  the  system  there  are  small  blunt 
protuberances  varying  from  a  few  millimeters  to  several  centimetres  long,  repre- 
senting abandoned  or  unfinished  tunnels.  Even  smaller  knobs  are  also  present. 
Where  several  branches  occur  close  together  the  tunnels  may  widen  to  form  a  flat 
chamber  (PL  8,  fig.  5)  up  to  10  cm.  long  and  5  cm.  wide  with  five  or  six  tunnels  leading 
off. 

The  most  striking  feature  of  the  branching  pattern  is  that  the  Y-forks  occur  in 
three  dimensions,  whilst  most  of  the  elements  of  the  system  are  horizontal,  joining 
into  small  irregular  polygons  (Text-figs.  4,  5,  a-b).  An  individual  system  can  exist 
at  several  levels,  connected  by  short  vertical  shafts.  At  Hunstanton,  these  levels 
are  5-6  cm.  apart,  running  along  the  minor  erosion  surfaces  within  the  Paradoxica 
bed,  although  elsewhere  levels  are  up  to  30  cm.  apart  (Text-fig.  4,  a).     The  systems 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 


143 


Fig.  4.  Thalassinoides  paradoxica  (Woodward),  a.  Junction  between  Wilmington  Sands 
and  the  overlying  Middle  Chalk;  White  Hart  Sandpit,  Wilmington,  S.  Devon.  Vertical 
Section,  x  \.  b.  Lower  Chalk,  Lower  Cenomanian,  Paradoxica  bed;  Hunstanton, 
Norfolk.     Plan,   x  i. 


are  connected  to  the  surface  by  vertical  or  steeply  inclined  shafts  15-30  cm.  long. 
On  vertical  faces,  burrow  densities  are  up  to  20/1000  cm.2 

As  pointed  out  by  Hughes  (1884)  the  surfaces  of  burrows  have  a  green  flakey  coat ; 
this  appears  to  be  glauconite.  When  cleaned,  the  surfaces  of  burrows  are  seen  to  be 
covered  with  rather  poorly  defined  longitudinal  ridges  (PI.  9,  fig.  2). 


i44  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Discussion.  The  name  Spongia  paradoxica  was  introduced  by  Woodward  in 
1830  in  a  "  Synoptic  table  of  British  Organic  remains  ".  In  illustration,  he  referred 
to  a  figure  in  a  paper  by  Webster  (1814)  "  On  some  new  variety  of  fossil  Alcyonia  ", 
recording  it  from  Hunstanton  and  Southbourne.  Subsequently  (1833)  Woodward 
recorded  this  form  from  the  Red  Chalk  and  Paradoxica  bed.  The  figure  referred  to 
(Webster  1814,  pi.  27,  fig.  1)  is  in  illustration  of  (p.  377)  "  an  account  of  the  singular 
organic  body  which  I  observed  in  the  green  sandstone  stratum  under  the  chalk  " 
of  the  Isle  of  Wight.  The  figure  is  of  a  large  block  of  Upper  Greensand  from  the 
Undercliff  along  the  southern  coast  of  the  island,  present  whereabouts  of  this  specimen 
unknown  !  Since  Woodward  records  Spongia  paradoxica  from  both  the  Red  Chalk 
and  Paradoxica  bed  at  Hunstanton,  other  material,  available  for  lectotype  designa- 
tion was  clearly  in  Woodward's  possession.  His  collection  (or  what  remained  of  it) 
passed  to  the  Norwich  Castle  Museum  in  1836,  and  was  incorporated  in  the  collections. 
Much  of  the  material,  including  figured  specimens,  was  missing  at  the  turn  of  the 
century,  and  although  Mr.  B.  McWilliams  of  the  Museum  has  searched  the  collections 
for  any  specimens  of  S.  paradoxica  from  Woodward's  collection,  or  labelled  in  his 
hand,  none  now  remains  which  can  be  attributed  to  him.  As  a  result,  I  have  selected 
a  specimen  from  the  Paradoxica  bed  at  Hunstanton  as  neotype. 

Although  first  named  by  Woodward,  these  bodies  were  noted  in  1823,  in  what  must 
be  one  of  the  earliest  records  of  a  burrow  from  the  chalk  (albeit  misidentified) . 
Taylor  (1823),  describing  the  sequence  in  the  cliffs  at  Hunstanton,  Norfolk,  noted  as 
follows : 

"  No.  4.  1^  feet.  A  stratum  of  white  chalk,  more  loose  than  the  last,  containing 
no  fossil  shells:  yet  it  is  to  be  distinguished  by  a  remarkable  ramifying  zoophyte, 
resembling  the  roots  of  trees;  about  an  inch  thick,  branching  and  intertwining  in 
every  direction.     Some  of  the  fragments  are  not  unlike  the  horns  of  a  stag." 

This  is  the  Paradoxica  bed. 
Again,  later: 

"  No.  6.  2  feet.  Red  Chalk,  of  a  rough  disjointed  structure,  similar  except  in 
colour  to  No.  4,  and  like  it,  though  in  a  smaller  degree,  interwoven  with  the  ramifying 
zoophytes  before  mentioned." 

The  first  use  of  the  name  paradoxica  was  by  Woodward  (1830 :  5) : 
"Spongia    paradoxica.      Geol.  Trans,    ii.    t.27,  f.i.      Red    Chalk.      Southbourn ; 
Hunstanton." 

The  Southbourne  occurrence  would  appear  to  be  the  same  as  that  given  by  Mantell 
(1833),  in  a  list  of  "  Fossils  from  the  chalk  formation  ",  where  a  spongia  from  South- 
bourne (Sussex)  is  noted.  In  a  footnote  stating,  "  the  inferior  bed  of  marl  which  is  in 
contact  with  the  Firestone  at  Southbourne  is  almost  entirely  composed  of  zoophytes, 
milleporites,  madreaporites  etc.,  so  as  to  form  coral  reefs  " — presumably  the 
Glauconitic  Marl. 

The  specimens  figured  and  described  by  Webster  (1814)  include  a  number  of  forms, 
both  burrows  and  fossil  sponges.  Plates  27  and  29  represent  burrows,  plate  28, 
figures  3  (in  part),  4,  8,  9,  10  and  11  represent  the  "  tulip  alcyonidium  "  (Siphonia 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  145 

tulipa  Zittel).  The  nature  of  the  specimens  figured  as  pi.  28,  figs.  5-7  is  not  clear, 
but  they  resemble  Cylindrites  spongioides  (Goeppert  1842),  here  regarded  as  a  crus- 
tacean burrow. 

The  next  reference  to  "  Spongia  paradoxica  "  is  by  Woodward  (1833  :   29) : 
"  Chalke  Marie.  This  bed  reposes  upon  the  red  chalk,  and  is  seen  to  great  advantage 
in  that  interesting  section,  Hunstanton  cliff.     It  is  of  a  grayish  color,  and  at  that  place 
about  four  feet  in  thickness.     The  Spongia  paradoxica,  as  we  have  named  it  pro 
tempore,  abounds  in  it  "  .  .  . 

Again,  (p.  30) : 

"  The  Red  Chalk  .  .  .  it  is  about  two  feet  in  thickness,  and,  like  its  superincumbent 
bed,  abounds  with  Spongia  paradoxica." 

Rose  (1835),  describing  the  Red  Chalk  at  Hunstanton  speaks  of  "  a  ramose  zoo- 
phite,  the  nature  of  which  is  not  satisfactorily  determined  ".  Again  (p.  275), 
under  the  heading  "  Chalk  without  flints  ",  describes  the  lowest  bed  as  being  "  made 
up  of  a  ramose  zoophyte,  which  strongly  characterizes  it  ",  regarding  it  (p.  276)  as 
"  best  explained  by  supposing  it  originally  a  coral  reef  and  its  interstices  filled  with 
Cretaceous  Mud." 

Wiltshire  (1859:  275),  in  a  list  of  fossils  from  the  Red  Chalk  records"  Spongia 
paradoxica  Geol.  Trans.  2,  tab.  27,  fig.  1.  page  377  (In  the  collections  of  Mr.  Rose 
and  Author.)  ".  Later  (p.  277,  footnote):  "  Siphonia  pyriformis  is  probably  the 
head  of  Spongia  paradoxica.  In  the  cabinet  of  Mr.  Rose  is  a  mass  of  the  latter,  to 
which  a  head  similar  to  the  one  figured  is  attached".  The  figures  referred  to  are  the 
first  of  S.  paradoxica  from  Hunstanton  published.  Wiltshire's  pi.  1,  fig.  1,  shows  a 
typical  branching  fragment,  fig.  2,  referred  to  as  Siphonia  pyriformis  is  a  swollen 
cylindrical  body,  the  terminal  portion  being  flat,  with  a  circular  depression  surrounded 
by  a  raised  rim. 

Seeley  (1864),  refers  to  "  organic  growth  known  as  '  spongia  paradoxica  '  in  the 
sponge  bed  and  top  of  the  red  chalk  ".  Wiltshire  (1869),  describing  the  Hunstanton 
section,  notes  "  a  meandering  and  many-branched  sponge,  Siphonia  paradoxica  " 
from  his  bed  b  (the  Paradoxica  bed),  whilst  "Spongia  paradoxica  Webster",  is 
recorded  from  the  highest  band  of  the  Red  Chalk. 

By  far  the  most  extensive  discussion  is  that  of  Hughes  (1884)  who  concludes 
(257-277)  that  Webster's  figure  is  a  different  fossil,  the  "  tulip  alcyonidium  ",and 
that  the  fossil  Spongia  paradoxica  is  in  fact  an  inorganic  body,  as  sponge  structure  is 
preserved  in  the  surrounding  rock,  but  never  in  5.  paradoxica.  Large  shell  fragments 
in  the  matrix  indicate  conditions  unsuitable  for  a  delicate  sponge,  fragments  of  the 
Spongia  are  never  found  in  the  matrix,  whilst  shell  fragments  are  avoided  and 
never  encrusted  as  would  be  expected  in  the  case  of  a  sponge. 

This  inorganic  origin  is  accepted  by  Whitaker  &  Jukes-Browne  (1899)  who  repeat 
Hughes'  views,  and  Jukes-Browne  (1900,  1903),  who  refers  to  "  curious  cylindrical 
bodies  .  .  .  which  resemble  the  stems  of  Siphonia  but  which  do  not  contain  any 
sponge  structure  ".  The  most  recent  account  of  this  "  organism  "  is  that  of  Peake  & 
Hancock  (1961),  describing  the  Paradoxica  bed: 


146 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 


B 


Fig.  5.  a,  b.  Thalassinoides  paradoxica  (Woodward),  a,  Lower  Chalk,  Lower  Ceno- 
manian,  Paradoxica  bed;  Hunstanton,  Norfolk.  Plan,  x  \.  b,  Middle  Chalk,  Lower 
Turonian,  Melbourn  Rock;  Brockham  Limeworks,  Betchworth,  Surrey.  Plan,  xj. 
c.  Thalassinoides  sp.  Lower  Chalk,  Middle  Cenomanian;  cliffs  below  Whitbred  Hole, 
Eastbourne,  Sussex.     Vertical  section  showing  entrance  shaft,   X  \. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  147 

"  soft-bodied  organisms  have  left  casts  of  their  burrows  which  now  form  radiating, 
cylindrical  branching  and  intertwining  masses ;  they  were  once  thought  to  be  sponges, 
and  actually  named  Spongia  paradoxica  " . 

Hantzschel  (1962)  includes  "  Spongia  paradoxica  Woodward  1833  (=Siphonia 
paradoxica  AUCTT.)  "  in  a  list  of  unrecognized  and  unrecognizable  genera  of  trace 
fossils,  giving  Taylor's  description  and  Hughes'  views  on  its  inorganic  origin. 

I  have  no  doubt  that  this  is  a  Thalassinoides,  from  both  mode  of  branching  and 
scratches;  and  that  it  is  a  crustacean  burrow.  The  irregular  three-dimensional 
branching  makes  it  readily  separable  from  other  forms  of  Thalassinoides.  The  short 
blind  burrows  give  a  superficial  resemblance  of  some  parts  to  Spongeliomorpha. 

Apart  from  its  record  in  the  English  literature,  already  discussed,  this  type  of 
burrow  was  described  as  early  as  1760  by  Schulze,  who  regarded  them  as  crinoid 
remains.  Problematic  structures  figured  by  Geinitz  (1871)  from  the  German  Upper 
Cretaceous  are  clearly  T.  paradoxica.  Spongia  sudolica  Zareczny  (1878),  regarded  by 
Raciborski  (1890)  as  a  Spongeliomorpha,  resembles  T.  paradoxica  in  size,  but  has 
rather  different  ornament;  it  is  clearly  a  Thalassinoides. 

In  Britain,  T.  paradoxica  has  a  very  limited  distribution,  occurring  only  in  associa- 
tion with  minor  erosion  surfaces  and  signs  of  early  lithification-hardgrounds.  At 
Hunstanton  the  burrows  occur  associated  with  the  erosion  surfaces  at  the  top  of  the 
Paradoxica  bed,  the  Red  Chalk  and  a  minor  erosion  surface  within  the  Inoceramus 
bed.  The  burrows  in  the  Paradoxica  bed  tend  to  spread  out  along  the  minor  erosion 
surfaces  within  the  bed,  but  often  pass  through  them.  At  the  base,  they  follow  the 
undulating  irregular  surface  of  the  Red  Chalk,  but  never  pass  into  it.  Those  in  the 
overlying  Inoceramus  bed  follow  the  surface  of  the  Paradoxica  bed  in  like  fashion. 
In  both  cases,  the  burrows  always  avoid  pebbles,  large  shell  fragments  and  echinoid 
tests,  indicating  that  the  animals  could  not  bore  into  hard  objects.  That  the  erosion 
surfaces  at  the  top  of  the  Red  Chalk  and  Paradoxica  bed  are  never  penetrated,  like- 
wise indicates  that  these  were  lithified  when  the  burrowers  were  active  in  the  sediment 
above.  In  the  burrows  beneath  the  Chalk  Rock,  here  preserved  as  empty  cavities, 
brachiopods,  echinoids  and  Inoceramus  fragments  protrude  into  Thalassinoides 
burrows,  the  surrounding  sediment  having  been  removed,  whilst  the  hard  shell 
was  left,  again  indicating  inability  to  deal  with  hard  objects. 

In  the  south,  the  top  of  bed  B.  in  the  Wilmington  outlier  is  penetrated  by  T. 
paradoxica  (Text-fig.  4,  a),  here  to  a  much  greater  depth,  as  there  is  no  lower  hard- 
ground  to  limit  penetration.  The  burrows  are  excavated  in  sandstone,  and  are  full 
of  glauconitic  chalk  with  small  phosphates,  identical  with  that  in  the  base  of  the 
overlying  Middle  Chalk  (equivalent  to  bed  C.  of  the  coastal  sections).  The  walls 
of  these  burrows  have  a  phosphatic  veneer,  as  does  the  overlying  erosion  surface, 
whilst  the  sediment  immediately  around  the  burrows  is  impregnated  with  glauconite ; 
this  suggests  that  the  lithification  prior  to  these  processes  occurred  either  whilst  the 
burrows  were  still  occupied,  or  in  the  period  before  they  were  filled  by  drifted  material. 

This  same  trace-fossil  occurs,  associated  with  hardgrounds,  at  the  top  of  beds  A2 
and  B  of  the  Cenomanian  Limestone  on  the  Devon  coast.  It  is  also  present,  asso- 
ciated with  hardgrounds,  in  the  Middle  Chalk  above,  and  elsewhere  in  Southern 
England  in  the  Melbourn  Rock,  again  associated  with  hardgrounds.     In  the  Lower 


148  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Chalk,  it  occurs  below  a  hardground  n  m.  below  the  plenus  Marls  at  Culver  Cliff 
(Isle  of  Wight). 

In  every  case  there  is  evidence  that  the  burrow  was  excavated  in  soft  sediment, 
and  that  the  hardening,  phosphatization  and  glauconitization,  some  or  all  of  which 
occur  in  the  associated  hardground,  post-date  burrowing,  but  seem  to  have  occurred 
prior  to  their  silting  up. 

I  have  never  seen  any  comparable  burrow  systems  in  chalk  away  from  hard- 
grounds,  and  believe  the  association  to  be  a  valid  one. 

The  "  tubulures  "  of  many  continental  authors,  occurring  associated  with  hard- 
grounds,  variously  interpreted  as  tree-roots,  algae,  annelid,  terebellid  or  crustacean 
burrows  (Schroeder  &  Bohm  1909,  Ellenberger  1946,  1947,  Marliere  1933  etc.),  are 
clearly  burrows  of  a  similar  type. 

Occurrence.  The  distribution  and  occurrence  of  T.  paradoxica  is  fully  dealt 
with  in  the  discussion  above. 


Ichnogenus  CHONDRITES  Sternberg  1833 

Diagnosis.  "  Very  plant-like,  regularly  ramifying  tunnel  structures  which 
neither  cross  each  other  nor  anastomose ;  should  be  interpreted  as  dwelling  burrows 
or  feeding  burrows;  width  of  tunnels  remaining  equal  within  a  system,  otherwise 
varying  from  large  (i.e.  Buthrotrephis)  to  small  (e.g.  Chondrites)  very  common  trace 
fossil,  usually  named  fucoid  .  .  .  surface  pattern  commonly  very  regular,  effected  by 
phobatactis  .  .  .  (probably  made  by  marine  worms).  Cambrian  to  Tertiary.  Cosmo- 
politan "  (Hantzschel  1962  :  187-188). 

Type  species.  Fucoides  targionii  Brongniart,  by  the  subsequent  designation  of 
Andrews  (1955). 

Discussion.  The  synonymy  of  Chondrites  is  given  by  Hantzschel  (1962).  Scott 
Simpson  (1957)  has  discussed  this  trace  fossil  at  length,  reviewing  early  interpretations 
and  concluding  that  it  is  the  feeding-trace  of  some  worm-like  organism. 

The  following  features  have  been  indicated  as  diagnostic  (Simpson  1957) : 

(a)  Circular  cross-section. 

(b)  Constant  diameter  (in  some  cases  with  constrictions  at  the  point  of  branching) . 

(c)  Smooth  wall. 

(d)  Regular  branching  pattern : 

(i)  Branching  tends  to  be  pinnate,  especially  at  the  periphery  of  the  system, 

when  not  interfered  with  by  neighbouring  systems, 
(ii)  Branching  is  always  lateral,  never  equal, 
(iii)  A  large  number  of  orders  of  branching  may  be  present. 
(iv)  The  pattern  lacks  symmetry  other  than  a  radial  tendency. 

(e)  Attitude,  with  both  vertical  and  horizontal  elements,  the  latter  undergoing 

extensive  ramification. 

(f )  Phobotactic  pattern. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  149 

Chondrites  sp. 

(PL  2,  figs.  2,^;   PL  5,  fig.  3;   PL  9,  fig.  1) 

A  small  species  of  Chondrites  with  tunnel  diameters  between  1  and  2  mm.  is 
common  throughout  the  whole  of  the  Lower  Chalk. 

Many  of  the  small,  horizontal  or  gently  inclined  burrows  present  can  be  referred 
to  this  genus;  every  section  I  have  examined  contains  this  form,  which  is  also  occa- 
sionally encountered  in  the  Glauconitic  Marl. 

Horizontal  sections  show  that  the  vertical  elements  of  these  systems  have  a  circular 
cross  section ;  vertical  sections  show  that  the  horizontal  or  gently  inclined  elements 
have  an  elliptical  section,  presumably  as  a  result  of  compaction  (which  can  be 
demonstrated  by  the  crushing  and  deformation  of  associated  fossils).  Diameters  are 
very  constant,  varying  between  1  and  2  mm.  in  specimens  from  many  horizons  and 
localities.  Individual  branches  and  systems  show  a  constant  diameter  throughout. 
Tunnel  walls  are  smooth,  tunnel  fills  structureless.  Tunnels  are  always  straight, 
except  at  the  point  of  branching. 

Vertical  elements  are  less  abundant  than  horizontal  or  inclined  parts.  No  example 
of  the  two  joining  up  has  been  observed  but  a  sharp  change  of  direction  is  implied. 
Sections  give  only  limited  information  about  branching  but  show  this  to  have  been 
lateral,  never  equal  and  at  an  acute  angle.  Horizontal  and  inclined  elements  branch 
frequently,  vertical  elements  rarely. 

Burrows  never  intersect,  suggesting  a  phobotactic  behaviour  pattern :  more  posi- 
tive evidence  is  seen  in  sections  which  suggest  a  "  wrapping  around  "  of  tunnels, 
embracing  on  close  encounter,  then  continuing  in  the  original  direction.  "  Solid  " 
specimens — chance  fracture  surfaces  (PL  9,  fig.  1)  or  the  cleaned  surfaces  of  larger 
burrows  from  beneath  the  Totternhoe  Stone  (PL  5,  fig.  3)  give  a  better  picture  of 
the  mode  of  branching.  All  the  features  already  noted  are  present.  There  is  no 
obvious  symmetry;  first  and  second  order  branches  are  present,  branching  at  acute 
angles;  pinnate  branching  is  occasionally  seen.  Phobotactis  is  expressed  in  terms  of 
"  embracing  "  and  stopping  short.  From  these  features,  reference  to  Chondrites  is 
clearly  justified. 

There  are  a  wide  variety  of  names  available  for  forms  of  this  size  and  it  seems  point- 
less to  name  the  present  material. 

As  noted  above,  every  section  contains  these  burrows:  the  maximum  observed 
density  is  about  20  sections  per  square  centimetre.  A  very  characteristic  occurrence 
of  Chondrites  is  in  the  filling  of  larger  burrows  (PL  2,  fig.  4) .  In  nearly  every  instance, 
Chondrites  is  far  more  abundant  in  these  than  in  the  surrounding  sediment ;  Thalas- 
sinoides  is  particularly  prone  to  this  re-working.  Beneath  the  Totternhoe  Stone  at 
Houghton  Regis,  the  filling  of  T.  saxonicus  burrows  is  completely  re-worked  and  the 
bottom  surface  of  the  burrow  converted  to  a  felted  mass  of  Chondrites  (PL  5,  fig.  3), 
whereas  this  form  is  uncommon  at  the  base  of  the  Totternhoe  Stone  and  penetrates 
only  a  few  centimetres.  Chondrites  penetrates  to  a  much  greater  depth  in  burrow 
fillings  than  the  surrounding  sediment;  a  similar  feature  has  been  noted  by  Seilacher 
(1964  :  302,  text-fig.  3,  right-hand  figure),  Chondrites  penetrating  to  a  greater  depth 


150  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

in  the  septum  of  a  Corophioides  than  in  the  surrounding  sediment  (Lias  y,  S.  Germany) . 
A  number  of  alternative  explanations  can  be  offered: 

(i)  The  filling  of  the  larger  burrows  is  richer  in  nutrients  than  the  surrounding 

sediment, 
(ii)  Re-worked  sediment  is  better  oxygenated, 
(iii)  Re-worked  sediment  is  softer  and  thus  more  readily  penetrated. 

Of  these  alternatives,  I  would  favour  (iii)  in  view  of  the  occurrences  beneath  the 
Totternhoe  Stone  where  the  presence  of  other  burrowers,  feeding  on  sediment,  suggests 
it  contained  nutrients  and  was  suitable  for  burrowing. 

The  abundance  of  Chondrites  on  the  lower  surface  of  larger  burrows  suggests  that 
there  is  a  geotropic  control  on  the  direction  of  burrowing,  and  that  the  sediment 
surrounding  the  burrows  is  not  penetrated  suggests  that  it  was  too  stiff  for  the 
Chondrites  animal.  The  surfaces  of  Gyrolithes,  as  figured  by  Saporta  (1884)  are 
covered  by  Chondrites  in  a  similar  fashion.  An  alternative  may  be  that  the  Chon- 
drites animal  was  feeding  on  mucus  lining  the  burrow. 

Ferguson  (1965)  has  suggested  that  the  filling  of  Chondrites  tunnels  was  by  the 
sucking-in  of  sediment  from  the  surface-opening  of  the  system  as  soon  as  the  proboscis 
(or  whatever  part  of  the  animal  produced  the  burrow)  was  withdrawn  from  a  branch. 
The  arguments  for  this  mode  of  filling  are  very  reasonable,  but  it  should  be  noted  that 
ammonite  siphuncles,  borings  and  echinoid  stereomes  are  sometimes  sediment  filled, 
indicating  that  passive  filling  of  such  structures  can  occur. 

Occurrence.  Chondrites  sp.  occurs  in  all  sections  of  Lower  Chalk  examined. 
Comparable  forms  occur  occasionally  in  the  Glauconitic  Marl  and  are  common  in  the 
plenus  Marls. 


Ichnogenus  SPONGELIOMORPHA  Saporta  1887 

1887  Spongeliomorpha  Saporta  :  299,  pi.  6,  figs.  2,  3. 

?igi3  Rhizocorallium;   Felix  :  21  (non  Zenker). 

1945  Spongiliomorpha;   Darder,  plate  8  (errore). 

1955  Spongeliomorpha]   de  Laubenfels  :  E  36. 

?i955  Felixium;   de  Laubenfels  :  E  36. 

1962  Spongeliomorpha;   Hantzschel  :  W  216. 

?ig65  Felixium;      Hantzschel  :  35. 

1965  Spongeliomorpha;   Hantzschel  :  87. 

1965  Spongiliomorpha;    Hantzschel  :  87. 

Diagnosis.  Medium  sized,  elongate,  cylindrical,  branching  tunnel  system, 
surfaces  covered  with  network  of  fine  ridges,  interpreted  as  scratch  marks ;  probably 
produced  by  crustaceans.     Range:  Triassic  to  Miocene. 

Type  species.  Spongeliomorpha  iberica  Saporta  1887  (299,  pi.  6,  figs.  2,  3) 
from  the  Miocene  of  Alcoy,  Spain,  by  monotypy. 

Discussion.  Saporta  (1887)  described  what  he  believed  to  be  a  new  form  of 
keratosid  sponge,  Spongeliomorpha  iberica,  based  on  material  from  the  Miocene  of 
Alcoy  (Spain),  comparing  it  with  Spongelia  Nardo  (in  fact  a  synonym  of  Dysidea 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  151 

Johnson:  see  de  Laubenfels  1955:  E  536),  a  form  ranging  from  Eocene  to  Recent. 
In  addition  to  the  type  material,  Saporta  mentions  other  material  from  the  Calcaire 
Grossier  and  the  United  States.  The  fossil  is  indicated  as  resembling  the  horns  of  a 
deer,  and  being  associated  with  Taonurus  Saporta  {=Rhizocor allium  Zenker),  a 
"  fucoid  ".  Meunier  (1889)  described,  without  figuring,  the  material  noted  from 
the  Calcaire  Grossier.  This  new  form,  Spongeliomorpha  saportai  Meunier,  from  the 
"  Sables  du  Beauchamp  ",  above  the  Calcaire  Grossier,  differs  from  5.  iberica 
in  its  more  elongate  form  and  tendency  to  dichotomous  branching  (the  specimen  is 
22  cm.  long  with  a  diameter  of  2  cm. ;  lateral  second  and  third  order  branches  are 
present).  The  surface  is  said  to  be  covered  by  ridges  more  regular,  parallel  and  uni- 
form than  in  S.  iberica.  Interpretation  of  Spongeliomorpha  as  a  sponge  is  supported 
by  Reis  (1910),  who  describes  Triassic  forms,  and  de  Laubenfels  (1955)  who  compares 
it  to  the  Jurrassic  form  Spongelites  Rothpletz,  a  genuine  sponge.  Darder  (1945) 
figures  "  Spongiliomorpha  "  iberica,  again  from  the  Miocene  (Burdigalian)  of  Alcoy, 
but  regards  it  as  algal,  and  a  sexual  dimorph  of  Taonurus  ultimus  (i.e.  a  Rhizo- 
corallium)  !  The  most  satisfactory  explanation  is  that  of  Reis  (1922)  who  interpreted 
Spongeliomorpha  as  a  burrow  system. 

The  genus  Felixium  de  Laubenfels  (1966) ,  proposed  to  replace  Rhizocorallium  Felix 
(1913,  non  Zenker)  with  R.  glaseli  Felix  (gldseli  recte  =  glaeseli)  as  type  species, 
appears  to  be  a  burrow,  perhaps  a  Spongeliomorpha,  perhaps  a  Thalassinoid.es  frag- 
ment or  even  the  "  arm  "  of  a  Rhizocorallium. 

Scratched  burrows  for  which  the  name  Spongeliomorpha  seems  suitable  have  been 
discussed  and  figured  by  Lessertisseur  (1955)  from  the  marine  Hauterivian  of  Andon 
(Alpes-Maritimes,  France)  and  Weigelt  (1929)  from  the  Jurassic  and  Cretaceous  of 
Germany.  Raciborski  (1890)  regarded  Spongia  sudolica  Zareczny  (1878)  as  a  Sponge- 
liomorpha; from  the  branching  pattern  it  is  clearly  a  Thalassinoides ,  possibly  a 
synonym  of  T.  paradoxica  (vide  p.  147). 

The  figured  material  of  Spongeliomorpha  is  all  in  the  form  of  small  fragments. 
The  original  figured  specimen  agrees  closely  with  some  fragments  of  Thalassinoides 
paradoxica  in  general  form,  whilst  the  ornament  of  Spongeliomorpha  and  Thalas- 
sinoides ornatus  suggests  that  when  the  branching  form  of  Spongeliomorpha  is 
better  known  the  two  names  may  prove  synonymous.  Spongeliomorpha  is  used 
in  the  present  account  for  scratched  burrows  which  do  not  show  a  Thalassinoides-like 
branching. 

The  surface  ridges  of  Spongeliomorpha  are  regarded  as  having  the  same  origin  as 
those  of  Thalassinoides — as  a  result  of  the  inhabitant  digging  or  moving  through 
the  system.  Once  more,  only  two  groups  of  animals  seem  likely  to  produce  these 
markings,  crustaceans  and  annelids.  Since  Weigelt  (1929)  has  figured  similar 
scratches  on  Recent  crustacean  burrows  whilst  the  same  ornamentation  is  present 
on  the  fossil  crustacean  burrow  Rhizocorallium  (Weigelt  1929,  Abel  1935,  Hantzschel 
1962),  a  crustacean  origin  for  Spongeliomorpha  is  clear.  Similar  ornamentation 
is  also  seen  on  the  undoubted  crustacean  burrow  Ophiomorpha  (personal 
observation  based  on  material  from  the  English  Weald  Clay  (Lower  Cretaceous, 
Barremian)). 


152  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Spongeliomorpha  sp. 

(PI-  7,  %•  7) 

Fragments  of  a  Spongeliomorpha  are  not  uncommon  in  the  Lower  Chalk  at  every 
locality  examined.  By  far  the  best  locality  is  beneath  the  Totternhoe  Stone  at 
Houghton  Regis.  The  cross  section  is  elliptical  (presumably  as  a  result  of  com- 
paction), varying  between  30  by  18  mm.  to  16  by  12  mm.  Straight  or  slightly  curved 
fragments  are  commonest  and  occur  in  both  vertical  and  horizontal  positions. 
Occasional  narrow  lateral  branches  may  be  present.  The  surface  is  covered  by  small, 
sharp  reticulate  ridges  intersecting  at  80  and  100  degrees.  No  internal  structure; 
surface  often  covered  with  small  Chondrites  and  other  burrows. 

In  size  and  general  form  these  fragments  are  closely  comparable  to  S.  iberica, 
differing  in  their  less  continuous  ridges  intersecting  at  a  higher  angle.  None  of  the 
material  I  have  seen  shows  the  antler-like  branching  of  the  figured  specimens. 

Occurrence.    Frequent  in  all  sections  of  the  Lower  Chalk  examined. 


Spongeliomorpha?  annulatum  ichnosp.  nov. 
(PI.  2,  fig.  1;   PI.  5,  fig.  5;   Text-fig.  6,  e) 

Diagnosis.  Cylindrical  branching  burrows  consisting  of  a  marl  cylinder  1-2-5  cm- 
in  diameter  with  a  glauconitic  core  5  mm.  in  diameter;  outer  surface  covered  by 
longitudinal  ridges.     Occurring  in  glauconitic  sediments. 

Holotype.  B.M.  (N.H.)  T.554  from  the  Glauconitic  Marl  (Lower  Cenomanian) ; 
section  below  the  Martello  Tower  No.  3,  Folkestone,  Kent. 

Material.  In  addition  to  the  holotype,  I  have  examined  many  hundreds  of 
specimens  from  the  Glauconitic  Marl  and  Upper  Greensand  of  Southern  England. 

Locality  and  horizon.  Abundant  in  the  Glauconitic  Marl  at  all  localities 
examined.  Occurring  also  in  glauconitic  bands  above  the  base  of  the  Chalk  and  in 
the  glauconitic  basement  bed  of  the  Lower  Chalk  in  the  south-west.  Very  common 
at  many  localities  and  horizons  in  the  Upper  Greensand.  Widespread  in  glauconitic 
facies  of  Cretaceous  age  all  over  north-west  Europe  (J.  M.  Hancock,  personal  com- 
munication). 

Description.  Largely  horizontal,  cylindrical  in  section  with  diameters  between 
i-o  and  2-5  cm.  Branching  poorly  known,  apparently  alternate  and  at  an  acute 
angle  (fig.  6,  e).  Occurring  only  in  glauconitic  sediments,  the  burrow  consists  of  a 
glauconite-free  marl  cylinder  with  a  central  glauconitic  core  about  5  mm.  in  diameter. 
The  outer  surface  of  the  marl  cylinder  is  covered  in  longitudinal  ridges. 

Discussion.  For  over  150  years  geologists  in  this  country  have  noted  the  presence 
of  "  stem-like  "  markings  in  the  Glauconitic  Marl  and  other  glauconitic  Albian  and 
Cenomanian  sediments.  Webster  (1814)  regarded  these  structures  as  alcyonites 
(sponges).  Reid  (1898),  describing  the  Upper  Greensand  near  Beachy  Head,  East- 
bourne (Sussex)  mentions  .  .  . 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  153 

"  curious  cylindrical  cavities  filled  with  material  differing  somewhat  from  the  sur- 
rounding matrix.  These  are  perhaps  made  by  some  boring  animal,  though  the 
horizontal  position  and  closed  ends  often  suggest  rather  the  disappearance  of  buried 
sand-eating  Holothurians." 

They  are  perhaps  the  "  irregular  spots  and  veinings  of  white  marl  "noted  by  Jukes- 
Browne  &  Hill  (1903  :  38)  from  the  Glauconitic  Marl  at  Folkestone  (Kent),  later 
(p.  265)  described  as  being  ..."  areas  of  small  size — seen  in  the  hand  specimen  as 
whitish  markings  or  pipings  are  filled  with  fine  amorphous  calcareous  material  to  the 
exclusion  of  the  larger  glauconite  grains  ".  What  are  probably  the  same  burrows 
are  noted  by  Thomel  (1961)  from  an  Upper  Albian  greensand  from  the  Alpes-Mari- 
times,  France. 

Although  extremely  abundant  in  most  sections  of  Glauconitic  Marl  and  Upper 
Greensand  (tunnel  densities  up  to  80  per  1,000  cm.2),  the  branching  pattern  is  poorly 
known.  The  surface  ornamentation  of  ridges  suggests  reference  to  Spongeliomorpha, 
but  because  of  the  peculiar  internal  structure  a  new  generic  name  may  be  useful  for 
this  type  of  burrow. 

Interpretation.  For  reasons  already  stated,  the  surface  ridges  of  these  burrows 
are  interpreted  as  scratch  marks  produced  by  crustaceans.  The  peculiar  internal 
structure  can  be  interpreted  as  a  result  of  the  sifting  of  sediment  into  clay,  silt  and 
sand  grade  materials  during  feeding,  the  animal  presumably  living  on  small 
organisms  in  the  coarse  fraction. 


Ichnogenus  PSEUDOBILOBITES  Lessertisseur  1955 

1882  Pseudobilobites,  Barrois  :  175,  pi.  5,  fig.  5a,  b  (not  intended  as  a  generic  name). 

1955  Pseudobilobite,  Lessertisseur,  text-fig.  25,  g. 

1955  Pseudobilobites  Barrois;   Lessertisseur  :  45. 

1965  ,,  Pseudobilobites  "  Barrois;  Hantzschel  :  75. 

Diagnosis.  Medium  sized  (3-7  cm.  long)  rounded  or  oval  masses  of  sand-grade 
microfossils  (largely  foraminifera)  and  shell  fragments  cemented  by  calcite,  generally 
ironstained,  due  to  oxidation  of  small  quantities  of  pyrite  present.  Upper  surface 
flat  or  concave,  smooth  or  slightly  granulated.  Lower  surface  convex,  convoluted, 
covered  by  groups  of  short  parallel  ridges  inclined  at  an  angle  to  the  axis  of  the  struc- 
ture. 

Type  species.  Pseudobilobites  jefferiesi  ichnosp.  nov.,  here  designated.  Lower 
Chalk,  Middle  Cenomanian;   Pitstone  (Bucks). 

Discussion.  The  term  "  pseudobilobite  "  was  first  used  by  Barrois  (1882),  in  a 
discussion  of  Bilobites  (=Cruziana) — resting  trails  of  trilobites,  from  the  Palaeozoic 
of  Northern  Spain.  Clearly  intended  as  a  vernacular  name,  he  applied  it  to  small 
oval  masses  of  microfossils,  the  lower  surfaces  of  which  are  covered  in  ridges,  from 
the  Lower  Turonian  of  Sery  in  the  Ardennes. 

Lessertisseur  (1955)  uses  the  term  rather  ambiguously :  in  the  explanation  of  his 
figure  25  G  (a  copy  of  Barrois  1882,  pi.  5,  fig.  5a)  he  uses  the  name  in  the  vernacular, 
on  page  45,  the  name  is  italicized,  as  are  the  other  generic  names  in  Lessertisseur, 


154  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

and  Barrois  is  given  as  author,  together  with  the  reference.  Clearly,  it  is  regarded 
as  of  generic  status.  As  already  noted,  Barrois  regarded  Pseudobilobites  as  a  ver- 
nacular name;  Lessertisseur's  use  as  a  generic  name,  with  Barrois  as  author  is  not 
justified.  The  genus  Pseudobilobites  is,  therefore,  attributed  to  Lessertisseur  1955. 
The  "  problematicum  "  of  Jefferies  (1962,  1963)  is  clearly  a  trace  fossil  of  this 
type.  Similar  forms  occur  in  the  Lower  Chalk  and  are  described  as  Pseudobilobites 
jefferiesi  ichnosp.  nov.,  here  designated  type  species  of  Pseudobilobites. 

Pseudobilobites  jefferiesi  ichnosp.  nov. 
(PL  6,  fig.  1;  PL  7,  fig  3;  PL  8,  fig.  4;  PL  9,  figs.  3,  4>  6) 

1961     Problematicum  sp.,  Jefferies,  text-fig.  2. 

1961     Problematicum,  Jefferies  :  620,  623,  624,  644,  pi.  77,  fig.  5. 

1963     Problematicum  sp.,  Jefferies  :  7,  12,  14,  16,  17,  text-fig.  2  (pars.). 

Diagnosis.    As  for  Genus. 

Holotype.  B.M.  (N.H.)  T.565.  Lower  Chalk,  Middle  Cenomanian;  Pitstone 
(Bucks.). 

Material.  Paratypes,  B.M.  (N.H.)  T.556,  566,  Lower  Chalk,  Upper  Cenomanian, 
10-15  ft.  below  base  of  plenus  Marls;  below  Shakespeare  Cliff,  Dover,  Kent. 
Numerous  other  specimens  from  the  Lower  Chalk  and  plenus  Marls  (Sedgwick 
Museum,  Cambridge,  Jefferies  collection). 

Description.  Small  ovoid  masses  of  sand-grade  microfossils  (foraminifera)  and 
shell  fragments,  cemented  by  crystalline  calcite  and  stained  brown  by  limonite, 
derived  from  the  decomposition  of  the  small  quantities  of  pyrite  present  in  un- 
weathered  specimens.  In  shape,  specimens  vary  from  elongate  ovals,  half  as  wide 
as  long,  to  almost  circular,  ranging  in  length  between  3  and  7  cm.,  although  larger 
specimens  probably  also  occur.  The  outline  is  fairly  regular,  although  often  broken 
up  by  subsequent  burrowing.  Upper  surface  smooth  or  slightly  granular,  flat  or 
slightly  concave,  lower  surface  convex,  convoluted  and  irregular,  covered  by  groups 
of  short,  parallel  ridges,  inclined  to  the  long  axis  of  the  structure. 

Discussion.  The  original  specimen  figured  by  Barrois  (1882)  differs  from  P. 
jefferiesi  in  having  longer  more  continuous  ridges  on  the  (presumed)  under-surface. 
The  figure  is  rather  indifferent  and  re-examination  of  the  material  may  indicate  that 
it  is  the  same  as  the  present  form. 

These  structures  were  first  recorded  from  the  English  Chalk  by  Jefferies  (1962, 1963) 
who  briefly  described  and  illustrated  a  "  problematicum  "  from  the  top  of  the  Lower 
Chalk  and  the  plenus  Marls.  Subsequent  collecting  shows  that  they  are  common 
throughout  the  whole  of  the  Lower  Chalk,  and  also  occur  in  the  Melbourn  Rock  at 
the  base  of  the  Middle  Chalk.  Specimens  show  great  shape  variation  in  both  outline, 
thickness  and  convexity,  but  form  a  quite  distinctive  group  of  trace  fossils. 

In  thin  section,  the  constituents  are  clearly  the  coarse  fraction  of  the  chalk. 
Foraminifera  are  abundant,  ostracods,  shell  and  echinoid  debris  plus  small  masses 
of  collophane  (?  faecal  pellets)  make  up  the  remainder,  with  a  calcific  cement. 
Burrows  passing  through  these  structures  suggest  they  were  soft  when  buried,  and 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  155 

that  they  are  not  of  diagenetic  origin  (PI.  9,  fig.  6).  Occasionally,  internal  lamina- 
tions are  present.  Pseudobilobites  clearly  represents  a  type  of  activity  like  that 
which  produced  what  I  have  called  "  laminated  structures  ".  The  prominent  ridges 
on  the  base  I  would  interpret  here,  as  elsewhere,  as  scratch  marks,  indicative  of 
crustaceans.  Grouping  in  threes,  fours  or  fives  represents  either  the  co-ordinated 
movement  of  appendages,  or  movement  of  a  single  appendage  with  several  claws. 
From  an  examination  of  these  structures  in  situ,  they  appear  to  be  a  surface  trace. 
I  have  never  seen  a  convincing  example  in  a  burrow,  although  the  possibility  cannot 
be  overlooked. 

This  type  of  structure  could  result  from  the  feeding  activities  of  an  animal  sifting 
chalk  for  the  fine  fraction,  ingesting  this  and  leaving  the  coarse  debris  behind.  The 
lower,  scratched  surface,  represents  the  extent  of  foraging,  the  concave  upper  surface 
is  perhaps  an  expression  of  the  position  of  the  body  during  feeding. 

Not  all  the  segregations  of  coarse  debris  in  the  Lower  Chalk  belong  to  this  form, 
some  (including,  in  part,  some  of  the  "  problematicum  "  recorded  by  Jefferies  (1961, 
1963))  represent  the  partial  or  total  filling  of  vertical  and  horizontal  cylindrical 
burrows  (a  typical  fragment  is  represented  in  PI.  6,  fig.  2).  This  type  of  filling 
probably  represents  the  same  type  of  activity.  They  sometimes  occur  closely 
associated  with  "  laminated  structures  "  (PL  8,  fig.  3)  and  may  be  the  product  of 
the  same  animal,  although  separate  occurrences  show  that  these  could  be  chance 
associations. 

P.  jefferiesi  is  widespread  and  common  in  the  Lower  Chalk  and  plenus  Marls,  also 

occurring  in  the  Melbourn  Rock  (Lower  Turonian). 


Keckia(?)  sp. 

191 1     Keckia  (?)  sp.,  Bather  :  553,  pi.  24,  fig.  1. 

Bather's  account  of  this  form  is  excellent,  as  is  his  illustration.  Having  seen  no 
other  material,  I  can  add  nothing  to  his  account.  The  nature  and  interpretation  of 
Keckia  has  been  discussed  by  Hantzschel  (1938)  and  by  Richter  (1947). 

"  Terebella  "  cancellata  Bather 
(PI.  8,  figs.  1,  2) 

1897     Terebella  lewesiensis  (Mantell)  Davies;    145-148  (pars.). 

191 1     "  Terebella  "  cancellata  Bather  :  551-553,  556,  pi.  24,  figs.  3,  4,  No.  5. 

Diagnosis.  "  Tube  from  which  the  (Pgelatinous  or  mucilaginous)  wall  has  dis- 
appeared, leaving  on  the  internal  cast  an  obscure  cancellate  ornament  formed  by 
transverse  and  longitudinal  folds;  with  diameters  from  about  0-75  to  2  cm.  and  with 
a  possible  length  of  19  cm.  or  more"  (Bather  1911). 

Holotype.  B.M.  (N.H.)  58253,  Lower  Chalk,  Glynde,  Sussex,  figured  here  as 
Plate  8,  fig.  1. 


156  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Discussion.  Bather's  description  of  this  "  Terebellid  "  is  excellent,  but  I  believe 
his  interpretation  to  be  erroneous.  "  Terebella  "  cancellata  is  clearly  a  burrow; 
material  agreeing  with  the  holotype  and  the  holotype  itself  all  appears  to  represent 
poorly  preserved  burrows  of  a  type  agreeing  with  what  I  have  called  Spongeliomorpha 
sp.  The  surface  depressions  described  by  Bather  (191 1)  are  the  result  of  rather 
poorly  preserved  intersecting  ridges  (i.e.  scratches).  The  paratype  specimen, 
B.M.  (N.H.)  1574  (PI.  8,  fig.  2)  clearly  belongs  to  a  different  form  and  is  described 
below,  as  burrow  type  D. 

From  a  re-examination  of  the  holotype  of  "  Terebella  "  harefieldensis  White 
(White  1923),  here  figured  for  the  first  time,  as  Plate  7,  fig.  2,  it  is  clearly  identical 
with  crustacean  "  burrows  "  figured  by  Weigelt  (1929),  from  a  similar  occurrence  in 
Germany.  T.  harefieldensis  is  not  a  true  burrow;  excavated  in  hard  chalk,  below 
the  sub-Tertiary  erosion  surface  it  is  to  be  regarded  as  a  boring.  From  its  widespread 
distribution  (Hester  1965,  text-fig.  2)  recognition  as  a  crustacean  boring  may  give 
this  form  value  as  a  palaeogeographic  indicator.  No  generic  name  appears  to  be 
available  for  this  type  of  boring. 


V.     OTHER  BURROWS 

The  forms  described  above  constitute  only  a  part  of  the  trace  fossil  assemblage  of 
the  Tower  Chalk.  Some  of  the  more  obvious  burrows,  too  poor  for  detailed  study 
are  noted  below. 

Burrow  Type  A 

(Text-fig.  6,  b,  c) 

Description.  Burrow  system  made  up  of  four  vertical  cylindrical  shafts  between 
6  and  12  cm.  long,  widening  downwards,  connected  by  a  horizontal  tunnel  16-30  cm. 
long.     Tunnel  diameters  about  2  cm. 

Discussion.  I  have  seen  only  two  complete  systems  of  this  type.  The  systems 
originate  at  the  bases  of  marls,  piping  down  into  the  limestones  below.  The  nature 
of  the  openings  is  not  clear,  but  the  vertical  shafts  increase  in  diameter  away  from 
the  surface  and  are  at  their  widest  just  above  the  junction  with  the  horizontal  tunnel. 
The  spacing  of  shafts  is  identical  in  both  examples  I  have  seen:  one  shaft  lies  at  each 
end,  the  other  two  are  equidistant  from  each  other,  but  one  is  separated  from  the  end 
by  nearly  twice  the  distance  separating  the  two  inner  shafts. 

I  have  seen  no  descriptions  or  figures  agreeing  with  these  systems,  and  in  view  of 
the  identical  form  of  the  two  examples,  I  am  inclined  to  regard  this  as  a  new  form. 

The  most  similar  described  system  is  that  of  Pholeus  abomasiformis  Fiege  (Fiege 
1944,  Hantzschel  1962,  1965)  from  the  Trias  (Muschelkalk)  of  North  Germany. 
Pholeus  differs  from  the  present  form  in  the  absence  of  intermediate  shafts  and  the 
presence  of  a  swollen  horizontal  chamber.  Pholeus  is  regarded  as  a  decapod  crusta- 
cean burrow.  The  burrow  of  the  living  crustacean  Cambarus  carolinus  Erichsen 
(Fiege  1944,  fig.  3)  is  again  similar  but  a  swollen  portion  ("  living  chamber  ")  is 
present. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 


157 


B 


Fig.  6.  a,  Burrow  type  B,  Top  of  Wilmington  Sands,  filled  by  the  overlying  Middle  Chalk; 
White  Hart  Sandpit,  Wilmington,  S.  Devon.  Vertical  section,  x  $■.  b,  c,  Burrow  type 
a,  Lower  Chalk,  Middle  Cenomanian.  b,  Folkestone,  Kent,  c,  Eastbourne,  Sussex.  Both 
vertical  sections,  x  J.  d,  Thalassinoides  sp.  Lower  Chalk,  Middle  Cenomanian ; 
Hunstanton,  Norfolk.  Plan,  showing  septate  internal  filling.  x  1.  e,  Spongeliomorpha? 
annulatum  ichnosp.  nov.     Upper  Greensand ;   Cow  Gap,  Eastbourne,  Sussex.     Plan,  xL 

GEOL.   15,  3.  16 


158  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

There  is  a  superficial  similarity  to  the  tube  system  of  the  living  polychaete  Lanice 
(Seilacher  1951,  Schaefer  1962)  but  the  Chalk  form  lacks  a  lining. 

This  system  cannot  be  compared  with  the  simple  U-shaped  burrow  of  worms  such 
as  Urechis  (MacGinitie  1928)  or  fossil  representatives  such  as  A renicolites,  as  the  pre- 
sence of  four  openings  would  render  the  functional  purpose  of  the  burrow,  in  terms 
of  maintenance  of  respiratory  and  feeding  currents,  too  complex. 

There  is  insufficient  evidence  to  suggest  the  nature  of  the  inhabitants  of  this  type 
of  system. 

Occurrence.  The  two  perfect  specimens  are  from  the  Middle  Cenomanian  of 
Folkestone  and  Eastbourne. 

Burrow  Type  B 

(Text-fig.  6a) 

Description.  Vertical,  cylindrical  burrows  up  to  5  cm.  in  diameter  and  100  cm. 
long.     Bottom  swollen  into  an  elongate  chamber. 

Discussion.  Large  simple  burrows  of  this  type  are  not  uncommon  in  the  upper 
part  of  the  sandy  facies  of  the  Cenomanian  at  the  White  Hart  Sandpit,  Wilmington, 
S.  Devon.  The  burrows  are  in  the  equivalent  of  bed  B.  of  the  coastal  sections, 
arising  from  the  erosion  surface  at  the  top  of  this  division  and  filled  with  the  overlying 
sandy  glauconitic  Middle  Chalk.  The  walls  of  these  burrows,  like  the  associated 
T.  pamdoxica  have  a  phosphatic  veneer  and  the  sediment  immediately  surrounding 
the  burrows  is  impregnated  with  glauconite. 

Similar  burrows  have  been  described  and  figured  by  Lessertisseur  (1955)  from  the 
Eocene  (Bartonian)  of  the  Paris  Basin  and  by  Maubeuge  &  Lanly  (1952)  from  the 
Bathonian  of  the  Vosges.  A  similar  but  much  smaller  form  occurs  in  the  Folkestone 
Beds  (Lower  Albian)  away  from  the  coastal  type  section,  as  at  Aylesford  (Kent), 
and  in  the  Woolwich  Bottom  Bed  (Eocene)  at  Upnor,  (Kent).  Some  of  the  specimens 
of  Cylindrites  spongioides  figured  by  Goeppert  (1842,  pi.  46,  figs.  1-4)  may  be  burrows 
of  this  type. 

These  burrows  are  very  similar  to  those  of  intertidal  crustaceans  from  the  East 
Indies  described  by  Verwey  (1930),  and  are  here  interpreted  as  the  work  of  crusta- 
ceans, although  it  is  not  implied  that  these  were  intertidal. 

Occurrence.  Top  of  Wilmington  Sands  (bed  B),  filled  with  Middle  Chalk; 
White  Hart  Sandpit,  Wilmington,  S.  Devon.  Similar  burrows  occur  in  the  bioclastic 
Santonian  of  the  Sudmerberg  near  Goslar  on  the  north  flank  of  the  Hartz  (J.  M. 
Hancock,  personal  communication). 

Burrow  Type  C 

(Text-fig.  7) 

Description.  Long,  straight  or  slightly  flexed,  very  narrow  cylindrical  burrows 
up  to  40  cm.  long  and  between  1  and  10  mm.  in  diameter.  Both  vertical  and  hori- 
zontal elements  occur,  the  latter  often  much  narrower  than  the  vertical  part,  from 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 


159 


J 


D 


Fig.  7.  Burrow  type  C.  a-d,  Lower  Chalk,  Middle  Cenomanian;  Glynde,  Sussex,  e-f, 
Lower  Chalk,  Upper  Cenomanian;  Dorking,  Surrey.  All  specimens  in  relief  on  vertical 
solution  planes.     All  x  j. 

geol.  15,  3.  i6§ 


160  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

which  they  branch  off  at  right  angles.  Some  of  these  burrows  curve  round  to  a 
horizontal  position  and  have  tunnels  10-25  cm-  l°ng- 

Discussion.  Burrows  of  this  type  are  best  seen  on  vertical  solution  planes, 
where  they  frequently  stand  out  in  relief.  From  their  abundance  on  such  surfaces 
they  are  clearly  responsible  for  much  of  the  sedimentary  mottling  seen  in  the  Lower 
Chalk. 

Some  fragments  compare  with  Lennea  Krausel  &  Weyland  (1932,  1934,  Paulus 
1957,  Hantzschel  1962,  1965)  from  the  Devonian  of  Germany,  but  are  generally 
smaller  and  lack  the  diagnostic  bifurcation  of  the  lateral  tunnels. 

It  would  seem  reasonable  to  regard  this  form  of  burrow  as  produced  by  worms  or 
some  worm-like  animal. 

Occurrence.  Present  in  all  sections  in  the  Lower  Chalk  examined,  particularly 
in  the  upper  part  where  they  are  prominent  on  vertical  solution  planes. 


Burrow  Type  D 
(PI.  5,  fig-  4;  PL  7.  figs-  1,  4,  5;  PL  9-  fig-  5) 

1897     Terebella  lewesiensis  (Mantell);    Davies  :  145-148  (pars.). 

191 1     "  Terebella  "  cancellata  Bather  :  551-553  (pars.),  pi.  24,  fig.  5  only. 

Description.  Vertical  and  horizontal  cylindrical  burrows,  straight  or  slightly 
sinuous,  unbranched  so  far  as  is  known,  between  5  and  25  mm.  in  diameter.  Frag- 
ments only  known,  up  to  20  cm.  long.  Surface  covered  in  long,  fine,  straight  or 
slightly  spiral  longitudinal  ridges,  also  bearing  coarser  longitudinal  folds. 

Discussion.  As  already  noted,  material  described  by  Bather  (1911)  as  "  Tere- 
bella "  cancellata  includes  two  distinct  forms.  The  forms  considered  here  are  those 
bearing  fine  longitudinal  ridges,  interpreted  by  Bather  (p.  552)  as  follows:  "  It  seems 
quite  certain  that  these  fossils  represent  tubes,  which  lay  on  the  sea-floor  or  in  the 
semi-floating  ooze  of  which  it  consisted,  and,  either  being  deserted  by  the  creature 
that  formed  them  or  persisting  after  its  death  and  decay,  were  filled  with  the  ooze  in 
which  they  lay.  The  tube  wall  it  is  clear,  was  of  such  strength  and  consistency  as  to 
retain  its  form  fairly  well  during  this  process,  and  yet  of  such  composition  that  it 
disappeared  after  the  partial  consolidation  of  the  ooze.  The  markings  on  the  infilling 
of  the  tube  may  be  due  to  two  causes ;  either  a  similar  folding  of  the  tube-wall  during 
life  or  a  wrinkling  and  contraction  of  the  tube  after  death  and  perhaps  even  after 
burial.  .  .  .  the  irregularity  and  variable  development  of  the  folds  suggest  that  they, 
at  least,  were  due  to  post-mortem  change  ". 

My  own  view  is  that  these  are  burrows,  and  that  they  never  existed  as  free  tubes 
on  the  surface. 

The  surface  features  of  these  burrows  can  be  interpreted  as  the  result  of  two 
different  processes.  The  fine  ridges,  I  would  interpret  as  the  result  of  some  worm-like 
animal  passing  through  sediment,  the  ridges  arising  from  bristles  or  appendages,  or 
even  grains  of  sediment  stuck  on  the  body.  The  longitudinal  folds  have  a  quite 
different  origin  and  appear  to  be  post-depositional  compaction  effects. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  161 

Clearly,  there  is  no  similarity  to  the  reticulate  surface  ornamentation  of  the  holo- 
type  of"  Terebella  "  cancellata. 

A  rather  similar  ornamentation  is  present  on  the  "  fucoids  "  Gyrolithes  dewalquei 
Saporta  (Saporta  1884)  Codites  neocomiensis  Saporta  &  Meunieur  (Saporta  1882) 
and  Cylindrites  rimosus  Heer  (Heer  1877). 

Occurrence.  Common  in  all  sections  examined  in  the  Weald,  Isle  of  Wight  and 
Chilterns. 

Burrow  Type  E 
(PI.  2,  fig.  3) 

Description.  Cylindrical  burrows,  generally  1-2  cm.  in  diameter,  known  only 
from  unbranched  fragments.  Sections  show  the  filling  of  these  burrows  is  septate, 
being  made  up  of  meniscus-shaped  laminae. 

Interpretation.  Burrow-fillings  showing  this  type  of  lamination  can  be  pro- 
duced by  a  number  of  groups.  Thalassinoides  occasionally  show  this  type  of  filling, 
as  do  other  undoubted  crustacean  burrows,  such  as  Ophiomorpha.  The  same  type 
of  structure  can  be  produced  by  coelenterates  (Schaefer  1962  :  326,  fig.  165), 
echinoids  (Schaefer  1962  :  348,  fig.  183)  and  some  bivalves  (Schaefer  :  424,  fig.  223). 
Under  the  conditions  of  chalk  sedimentation,  and  by  comparison  with  other  forms, 
these  are  probably  crustacean  burrows. 

Occurrence.     Uncommon  in  all  sections  of  the  Lower  Chalk  examined. 

Many  other  trace  fossils  are  represented  in  the  Lower  Chalk.  "  Terebella " 
lewesiensis  (Mantell),worm  tubes  lined  with  fish,  plant  or  echinoderm  debris  should  be 
interpreted  as  trace  fossils,  as  should  the  micro-coprolites  described  by  Wilcox 
(1953)  from  the  Upper  Chalk,  which  also  occur  in  the  Lower.  Borings,  in  shells, 
pebbles  and  rock  surfaces  are  very  abundant.  In  addition  to  species  of  Cliona, 
other  sponge  borings  (Filuroda) ,  algal  and  fungal  perforations  (Calcideleclrix,  Dictyo- 
porus),  cirripede  bores  (Zapfella,  Roger ella),  bryozoan  borings  and  bivalve  crypts  all 
occur. 


VI.    conclusions 

The  activities  of  burrowing  organisms  are  shown  to  be  universally  present  in  the 
Lower  Chalk.  The  most  obvious  are  those  of  crustaceans  (Thalassinoides ,  Spongelio- 
morpha)  and  "  worms "  (Chondrites).  Several  poorly  known  burrows  are  also 
described.  Of  previously  described  assemblages,  the  present  one  compares  best 
with  the  Lower  Lias  (Hallam  1961),  where  both  Chondrites  and  Thalassinoides  occur. 
U-shaped  burrows  (Rhizoco? ■allium  etc.),  common  in  the  Lower  Lias,  are,  however, 
absent  in  the  Chalk. 

A  problem  of  the  Lower  Chalk  fauna,  in  view  of  the  abundance  of  burrows,  is  the 
absence  or  great  rarity  of  the  animals  responsible.  With  worms,  disappearance  of 
the  soft  body  is  readily  understood,  but  the  absence  of  crustaceans  demands  explana- 
tion.    The  crustacean  fauna  of  the  Lower  Chalk  is  very  limited.  By  far  the  most 


162  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

abundant  form  is  the  large,  lobster-like  Enoploclytia,  though  the  very  massive 
claws  and  thick,  thorny  carapace  suggest  that  it  did  not  burrow.  The  only  other 
macrurous  crustacean  I  have  seen  is  Glyphea  willeti  (Woodward),  which,  in  view  of 
the  thin,  rather  delicate  exoskeleton,  could  well  have  burrowed.  In  size,  it  would 
fit  some  of  the  larger  Thalassinoides,  but  it  is  rare.  Callianassids,  recorded  in  associa- 
tion with  Thalassinoides  elsewhere,  appear  to  be  totally  absent  from  the  Lower 
Chalk  facies  of  the  Cenomanian,  although  a  "  Callianassa  "  sp.  is  present  in  division  A 
of  the  Cenomanian  Limestone  of  S.  Devon.  Hume  (1897)  records  a  Callianassa  sp. 
as  occurring  commonly  in  the  Upper  Glauconitic  Beds  (Cenomanian)  at  Colin  Glen, 
Co.  Antrim.  Callianassids  also  occur  in  the  Upper  Greensand  of  the  Devon  Coast 
(matrix  of  museum  specimens  suggests  the  Top  Sandstones)  and  the  Gault. 

Brachyurous  crustaceans  are  equally  rare;  a  few  specimens  of  Diaulax  and 
Necrocarcinus  are  known  from  the  Lower  Chalk,  whilst  crabs  are  not  uncommon  in 
the  sandy  facies  of  the  Cenomanian,  particularly  at  Wilmington.  In  all,  the  known 
crustacean  remains  give  few  clues  to  the  identity  of  the  burro wers.  A  possible 
explanation  of  absence  is  suggested  by  recent  burrowing  forms  which  have  a  thin, 
sometimes  even  transparent  exoskeleton,  very  poorly  calcified.  Sloughs  are  generally 
removed  from  burrows,  whilst  moribund  individuals  leave  their  burrows  prior  to 
death. 

Under  these  conditions  it  seems  possible  that  on  the  Lower  Chalk  sea  floor  the 
organically-rich  remains  were  completely  eaten  or  destroyed  by  scavengers  and 
micro-organisms  prior  to  burial. 

VII.     ACKNOWLEDGMENTS 

I  am  grateful  to  Dr.  J.  M.  Hancock  and  Dr.  J.  D.  Taylor  for  reading  the  manuscript 
of  this  paper  and  for  making  many  helpful  suggestions.  I  have  profited  from  dis- 
cussions with  Dr.  R.  Bromley,  Dr.  C.  V.  Jeans,  Dr.  R.  P.  S.  Jefferies  and  many  others; 
Mr.  P.  Palmer  kindly  provided  me  with  photographs  of  the  types  of  Terebella  cancellata 
and  T.  harefieldensis;  Mr.  R.  Cleevely  has  aided  me  greatly  in  finding  some  of  the 
more  obscure  literature.  Dr.  H.  W.  Ball,  Mr.  S.  Ware  and  Dr.  C.  L.  Forbes  have  kindly 
allowed  me  to  examine  material  in  their  care,  whilst  Dr.  Jefferies  has  generously 
allowed  me  to  work  on  his  collection,  now  in  the  Sedgwick  Museum.  Mr.  B. 
McWilliams  of  the  Norwich  Castle  Museum  searched  the  collections  for  type  material 
of  T.  paradoxica.  Preparation  of  illustrations  by  the  Technical  Staff  of  King's 
College,  under  the  direction  of  Mr.  E.  O.  Rowlands  is  acknowledged,  particularly  the 
assistance  of  Miss  M.  Baker  in  preparation  of  the  plates.  I  am  deeply  grateful  to 
my  parents  for  their  encouragement  and  assistance. 

Part  of  the  work  was  carried  out  under  the  tenure  of  a  N.E.R.C.  grant,  which  is 
gratefully  acknowledged. 

VIII.     REFERENCES 
Abel,  O.     1935.      Vorzeitliche  Lebensspuren.     644  pp.,  530  figs.     Jena. 
Andert,  H.     1934.     Die  Kreideablagerungen  zwischen  Elbe  und  Jeschken.     III.     Die  Fauna 

der  Obersten  Kreide  in  Sachsen,  Bohmen  und  Schlesien.     Abh.  preuss.  geol.  Landesanst., 

Berlin,  159  :  1-477,  pis.  1-2 1. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  163 

Andrews,  H.  N.     1955.     Index  of  Generic  Names  of  Fossil  Plants  1820-1950.  Bull.  U.S.  Geol. 

Sum.,  Washington,  1013  :  1-262. 
Arkell,  W.  J.     1935.     The  Portland  Beds  of  the  Dorset  Mainland.     Proc.  Geol.  Assoc,  London, 

46  :  301-347,  pis.  19-26. 
Barrois,  C.     1882.     Recherches  sur  les  terrains  anciens  des  Asturies  et  de  la  Galicie.     Mem. 

Soc.  Geol.  Nord,  Lille,  2  :  1-630,  pis.  1-20. 
Bather,  F.  A.     191 1.     Upper  Cretaceous  Terebelloids  from  England.     Geol.  Mag.,  London  (5) 

8  :  481-487,  549-556,  pi.  24. 
Billinghurst,  S.  A.     1927.     On  some  New  Ammonoidea  from  the  Chalk  Rock.     Geol.  Mag., 

London,  64  :  511-518,  pi.  16. 
Black,  M.     1953.     The  Constitution  of  the  Chalk.     Proc.  geol.  Soc.  Lond.,  1499  :  81-86. 
Black,  M.  &  Barnes,  B.     1959.     The  Structure  of  Coccoliths  from  the  English  Chalk.     Geol. 

Mag.,  London,  96  :  321-328,  pis.  8-12. 
Blake,  J.  F.  &  Hudleston,  W.  H.     1877.     On  the  Corallian  Rocks  of  England.     Quart.  Jl. 

geol.  Soc.  Lond.,  33  :  260—405,  pis.  12-17. 
Burnaby,  T.  P.     1962.     The  Palaeoecology  of  the  Foraminifera  of  the  Chalk  Marl.     Palaeon- 
tology, London,  4  :  599-608. 
Cayeux,   L.     1897.     Contribution  a  l'6tude  micrographique  des  terrains  s£dimentaires.     II. 

Craie  du  Bassin  de  Paris.     Mem.  Soc.  Geol.  Nord,  Lille,  4  :  206-563,  pis.  9,  10. 
Dales,  R.  P.     1957.     Interrelations  of  organisms.     A.     Commensalism.     In  Hedgepeth,  J.  W. 

(Ed.)  Treatise  on  Marine  Ecology  and  Palaeoecology,  1,  Ecology  Mem.  Geol.  Soc.  Amer., 

67  :  391-412,  pi.  1. 
Darder,  B.     1945.     Estudio  geologico  del  Sur  de  la  Provincia  de  Valencia  y  Norte  de  la  de 

Alicante.     Bol.  Inst.  Min.  Geol.  Espana,  Madrid,  57  :  59-362,  pis.   1-11. 
Davidson,  C.     1891.     On  the  amount  of  sand  brought  up  by  Lobworms  to  the  surface.     Geol. 

Mag.,  London  (3)  8  :  489-493. 
Davies,  W.     1879.     On  some  Fish  Exuviae  from  the  Chalk,  generally  referred  to  Dercetis 

elongatus,  Ag. :   and  on  a  New  Species  of  Fossil  Annelide,  Terebella  Lewesiensis.     Geol.  Mag., 

London  (2)  6  :  145-148. 
Dettmer,  F.     1912.     Spongites  saxonicus  Geinitz  und  die  Fucoidenfrage.     Neues  Jb.  Min.  Geol. 

Palaont.,  Stuttgart,  1912  :  1 14-126,  pis.  8,  9. 
1914.     Die  Spongites-saxonicus — Frage.     Sitz.-Ber.  u.  Abh.  naturwiss.   Gesellsch.   '  Isis  ' , 

Dresden,  1913  :  50-54. 
1915.     Neus  zum  Fucoidenproblem.     Centralbl.  Min.  Geol.  Pal.,  Stuttgart,  1915  :  285-287, 

1  fig. 
Ehrenberg,  K.     1938.     Bauten  von  Decapoden  (Callianassa  sp.)  aus  dem  Miozan  (Burdigal) 

bei  Eggenburg  im  Gau  Nieder-Donau  (Niederosterreich) .     Palaont.  Z.,  Berlin,  20  :  263-284, 

pis.  27-29. 
1944.     Erganzende  Bemerkungen  zu  den  seinerzeit  aus  dem  Miozan  von  Burgschleinitz 

beschriebenen  Gangkernen  und  Bauten  dekapoder  Krebse.     Palaont.  Z.,  Berlin,  23  :  354- 

359- 
Eichwald,  E.  d'     1865.     Lethaea  Rossica  2,  1,  Periode  Moyenne.     640  pp.     Stuttgart. 
Ellenberger,  F.     1946.     Observations  nouvelles  sur  la  craie  jaune  a  tubulures  de  Meudon. 

C.R.  Soc.  Geol.  Fr.,  Paris,  1946,  11  :  202-204. 
1947-     Le  probleme  lithologique  de  la  craie  de  Meudon.     Bull.  Soc.  Geol.  Fr.,  Paris  (5) 

17  :  255-274. 
Esper,  E.  J.  C.     1799.     Die  Pflanzenthiere.     In  Abbildungen  nach  der  Natur  mil  Farben  er- 

leuchtet  nebst  Beschreibungen,  2,  12  :  221-304,  pis.     Nuremberg. 
Farrow,  G.  E.     1966.     Bathymetric  zonation  of  Jurassic  trace  fossils  from  the  coast  of  York- 
shire, England.     Palaeogeography ,  Palaeoclimatol.,  Palaeoecol.,  2  :  103-151,  pis.   1-7. 
Felix,  J.     1913.     Uber  ein  cretaceisches  Geschiebe  mit  Rhizocor allium  Gldseli  n.  sp.  aus  dem 

Diluvium  bei  Leipzig.     Sitz.-Ber.  naturf.  Ges.  Leipzig,  39  :  19-25,  37,  pi.  1. 
Ferguson,  L.     1965.     A  note  on  the  emplacement  of  sediment  in  the  trace-fossil  Chondrites. 

Proc.  geol.  Soc.  Lond.,  1622  :  79-82,  fig.  1. 


1 64  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Fiege,    K.     1944.     Lebensspuren   aus   dem   Muschelkalk   Nordwestdeutschlands.     Neues  Jb. 

Min.,  Geol.,  Paldont.,  Stuttgart  (B)  88  :  401-426,  6  figs. 
Fisher,  W.  K.  &  MacGinitie,  G.  E.     1928.     The  natural  history  of  an  echiuroid  worm.     Ann. 

Mag.  Nat.  Hist.,  London  (10)  1  :  204-213,  pi.  1. 
Fric,  A.     1878.     Studien  im  Gebiete  der  bohmischen  Kreideformation.     II.     Die  Weissenberger 

und  Malnitzer  Schichten.     Arch.  naturw  .Landesdurchf .  Bohm.,  Prague,  4,  1  :  1-151. 
Geinitz,  H.  B.     1839-42.     Charakteristik  der  Schichten  und  Petrefacten  des  sachsisch-bohmischen 

Kreidegebirges.     116  pp.,   24  pis.     Dresden   &  Leipzig. 
1871-75.     Das  Elbthalgebirge  in  Sachsen.     Palaeontographica,  Stuttgart,  20  :  1-319,  pis. 

1-67. 
Gignoux,  M.     1926.     Geologie  Stratigraphique.     vi  +  588  pp.,   124  figs.     Paris. 
Glaessner,  M.  F.     1947.     Decapod  Crustacea  (Callianassidae)  from  the  Eocene  of  Victoria. 

Proc.  R.  Soc.   Vict.,  Melbourne,  59  :  1-7,  pis.  1,  2. 
Gmelin,  J.  F.     1793.     C.  a.  Linne  .  .  .  Sy sterna  Naturae  .  .  .  Editio  decima  tertia  aucta  reformata 

cura  J.  F.  Gmelin,  3.     476  pp.,  3  pis.     Lipsiae. 
Goeppert,  H.  R.     1842.     tJber  die  fossile  Flora  der  Quadersandsteinformation  in  Schlesien, 

als  erster  Beitrag  zur  Flora  der  Tertiargebilde.      Nova  Acta  Caes.  Leop. -Carol.,  Breslau  & 

Bonn,  19,  2  :  97-134,  pis.  46-53. 
1847.     Zur  Flora  des  Quadersandsteins  in  Schlesien  als  Nachtrag  zu  der  friiher  erschienenen 

Abhandlung  iiber  denselben  Gegenstand.     Nova  Acta  Caes.  Leop— Carol.,  Breslau  &  Bonn, 

22,  1  :  356-365,  Pis-  35-38- 
1854-     Quadersandsteinformation   insbesondere   iiber    Cylindrites   spongioides   m.      Verh. 

nat.  Ver.  preuss.  Rheinlande  u.  Westph.  (n.f.)  11  :  229-234,  pi.  3. 
Hallam,  A.     1961.     A  sedimentary  and  faunal  study  of  the  Blue  Lias  of  Dorset  and  Glamorgan. 

Phil.  Trans.  R.  Soc,  London  (B)  243  :  1-44,  pis. 
Hancock,  J.  M.     1963.     The  hardness  of  the  Irish  Chalk.     Irish  Nat.  J.,  Belfast,  14  :  157-164. 
Hantzschel,  W.     1934-     Schraubenformige  und  spiralige  Grabgange  in  turonen  Sandsteinen 

des  Zittauer  Gebirges.     Senckenbergiana,  Frankfurt  am  Main,  16  :  313-324,  4  figs. 
1935-     Xenohelix  saxonica  n.  sp.  und  ihre  Deutung.     Senckenbergiana,  Frankfurt  am  Main, 

17  :  105-108. 
1938.     Quer-Gliederung   bei   Littorina-Fahrten,    ein    Beitrag   zur   Deutung   von   Keckia 

annulata  Glocker.     Senckenbergiana,  Frankfurt  am  Main,  20  :  297-304,  6  figs. 
1952.     Die  Lebensspuren  von  Ophiomorpha  Lundgren  im  Miozan  bei  Hamburg,  ihre  welt- 

weite  Verbreitung  und  Synonymic    Mitt.  Geol.  Staats  inst.  Hamburg,  21  :  142-153,  pis.  13, 

14. 
1962.     Trace  fossils  and  Problematica.     Treatise  on  Invertebrate  Paleontology,  Part  W  : 

177-245,  figs.  109-149.     New  York. 
1964.     Spurenfossilien   und   Problematika  im   Campan   von   Beckum(  Westf.).   Fortschr. 

Geol.  Rheinl.  Westf.,  Krefeld,  7  :  295-308. 
1965.     Fossilium  Catalogus  1  :  Animalia,  Pars.   108,  Vestigia  Invertebratorum  et  Prob- 
lematica.    140  +  2  pp.  s'  Gravenhage. 
Heer,  O.     1877.     Flora  Fossilis  Helvetiae.     Die  vorweltliche  Flora  der  Schweiz.     182  pp.,  70  pis. 

Zurich. 
Hester,  S.  W.     1965.     Stratigraphy  and  Palaeogeography  of  the  Woolwich  and  Reading  beds. 

Bull.  Geol.  Surv.  G.B.,  London,  23  :  177-137,  pi.  10. 
Hughes,  T.  M.     1884.     On  the  so-called  Spongia  paradoxica  (S.  Woodward)  from  the  Red  and 

White  Chalk  at  Hunstanton.     Quart.  Jl.  geol.  Soc.  Lond.,  40  :  273-279. 
Hume,  W.  F.     1893.     Chemical  and  Micro-mineralogical  Researches  on  the   Upper  Cretaceous 

Zones  of  the  South  of  England.     103  pp.     London. 
1897.     The  Cretaceous  Strata  of  County  Antrim.     Quart.  Jl.  geol.  Soc.  Lond.,  53  :  540-606, 

pis.  44,  45. 
Jefferies,  R.  P.  S.     1962.     The  Palaeoecology  of  the  Actinocamax  plenus  subzone  (Lowest 

Turonian)  in  the  Anglo-Paris  Basin.     Palaeontology,  London,  4  :  609-647,  pis.  77-79. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  165 

Jefferies,  R.  P.  S.     1963.     The  Stratigraphy  of  the  A ctinocamax  plenus  subzone  (Turonian)  in 

the  Anglo-Paris  Basin.     Proc.  Geol.  Ass.,  London,  74  :  1-33,  pis.  1,  2. 
Jukes-Brown,  A.  J.  &  Hill,  W.     1900.     The  Cretaceous  Rocks  of  Britain.     Vol.1 — Gault  and 

Upper  Greensand.     Mem.  Geol.  Surv.  U.K.,  London,     xiv  +  499  pp.,  5  pis. 
1903.     The  Cretaceous   Rocks  of  Britain.     Vol.   2. — The  Lower  and  Middle   Chalk  of 

England.     Mem.  Geol.  Surv.  U.K.,  London,     xiii  +  556  pp.,  8  pis. 
1904.     The  Cretaceous  Rocks  of  Britain.     Vol.  3 — The  Upper  Chalk  of  England.     Mem. 

Geol.  Surv.   U.K.,  London,     x  +  566  pp.,  pi.  9. 
Keij,  A.  J.     1965.     Miocene  trace-fossils  from  Borneo.     Palaont.  Z.,  Stuttgart,  39  :  220-228, 

pis.  28,  29. 
Kennedy,  W.  J.     1967.  Field  meeting  at  Eastbourne,  Sussex.     Proc.  Geol.  Ass.,  London,  77  : 

365-370- 

Kilpper,  K.    1962.    Xenohelix  Mansfield  1927  aus  der  miozanen  niederrheinischen  Braunkohlen- 

formation.     Palaont.  Z.,  Stuttgart,  36  :  55-58,  pi.  7. 
Krausel,  R.  &  Weyland,  H.     1932.     Pfianzenreste  aus  dem  Devon,  II.     Senckenbergiana, 

Frankfurt  am  Main,  14  :  185-190. 
1934-     Lennea  schmidti,  eine  pflanzenahnliche  Tierspur  aus  dem  Devon.     Palaont.  Z., 

Berlin,  16  :  95-102,  pis.  7-1 1. 
Lambrecht,    F.     1927.     Schichtenfolge    und    Oberflachenformen    im    Winterberggebiete    des 

Elbsandsteingebirges.     Mitt.    Ver.   Erdk.,   Dresden    (N.F.)    1927  :  1-48. 
Laubenfels,  M.  W.  de     1955.     Porifera  :  Treatise  on  Invertebrate  Paleontology,  PartE  :  21-112, 

figs.  14-89.     New  York. 
Lessertisseur,   J.     1955.     Traces  fossiles  d'activit6  animale  et  leur  significance  pal6obiolo- 

gique.     Mem.  Soc.  Geol.  Fr.,  Paris  (n.s.)  74  :  1-150,  pis.  1-11. 
MacGinitie,  G.  E.     1930.     The  Natural  History  of  the  Mud  Shrimp  Upobegia  pugettensis  Dana. 

Ann.  Mag.  Nat.  Hist.,  London   (10)   6  :  36-44,  pis.    1-3. 
1934-     The    Natural    History   of    Callianassa   californiensis    Dana.     Amer.    Midi.    Nat., 

Notre  Dame  (Indiana),  15  :  166-177,  pls-  5-  6. 
Mantell,  G.  A.     1822.     The  fossils  of  the  South  Downs.     320  pp.,  43  pis.     London. 

1833.     The  Geology  of  the  South-East  of  England,     xix  +  viii  +  415  pp.,  7  pis.     London. 

Marliere,  R.     1933.     De  nombreux  bancs  phosphates  dans  la  craie  a  Actinocamax  quadratus 

du  basin  de  Mons.     Ann.  Soc.  geol.  Belg.,  Bruxelles,  56  :  289-301,  pis.  1,  2. 
Maubeuge,  P.  L.  &  Lanly,  R.     1952.     Sur  un  probtematica  du  Bathonien  moyen  de  la  region 

de  Neufchateau.     Bull.  Soc.  Sci.  Nancy,   11  :  122-124,   ng- 
Mertin,  H.     i  941.     Decapode  Krebse  aus  dem  subhercynen  und  Braunschweiger  Emscher  und 

Untersenon  sowie  Bemerkungen  fiber  einige  verwandte  Formen  in  der  Oberkreide.     Nova 

Acta  Leop. -Carol.,  Halle  (Saale),  10,  68  :  149-262,  pis.  1-8. 
Meunier,  S.     1889.     Sur  La  Spongeliomorpha  saportai  esp6ce  nouvelle  parisienne  C.R.  Acad. 

Sci.  Paris,  109  :  536-537. 
Moore,  H.  B.     1932.     Faecal  pellets  of  the  Anomuran  Crustacea.     Proc.  Roy.  Soc.  Edinb., 

52  :  296-308. 
1939.     Faecal  pellets  in  relation  to  Marine  deposits.     In  Trask,  P.  D.,  Recent  Marine 

Sediments,  A   symposium  :  516-524,  fig.   1.     Tulsa. 
Oosting,  C.  H.     1925.     Report  on  a  collection  of  recent  shells  from  Obi  and  Halmaher  (Mol- 

lucas).     Meded.  Landb.  Hoogensch.  Wageningen.,  29  :  1-363. 
Otto,  E.  v.     1852.     Additamente  zur  Flora  des  Quadergebirges  in  der  Gegend  um  Dresden  und 

Dippoldiswalde,  enthaltend  meist  noch  nicht  oder  wenig  bekannte  fossile  Pflanzen.     29  pp., 

12  pis.     Dippoldiswalde. 
Paulus,  B.     1957.     Das  Spurenfossil  Lennea  schmidti  im  Devon  der  Eifel.     Senckenberg.  leth., 

Frankfurt  am  Main,  38  :  169-175,  pi.  1. 
Peake,  N.  B.  &  Hancock,  J.  M.     1961.     The  Upper  Cretaceous  of  Norfolk.     Trans.  Norfolk 

Norw.  Nat.  Soc,  Norwich,  19  :  293-339,  pi.   1. 
Pocta,    P.     1885.     Beitrage   zur   Kenntnis   der   Spongien   der   bohmischen    Kreideformation. 

Abh.  bohm.  Ges.  Wiss.,  Prague,  3  :  1-40. 


166  LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND 

Pohl,  M.  E.     1946.     Ecological  Observations  on  Callianassa  major  Say  at  Beaufort,  North 

Carolina.     Ecology,  Brooklyn,  27  :  71-80. 
Prescher,  H.     1954.     Sedimentpetrographische  Untersuchungen  oberturoner  Sandsteine  im 

Elbsandsteingebirge.     Freiberger  Forschungsh.,  C  11  :  1-96,  28  figs. 
Pruvost,  P.  &  Pringle,  J.     1924.     Excursion  to  the  Boulonnais.     Proc.  Geol.  Ass.,  London, 

35  :  56-67,  pis.  5-7. 
Purdy,  E.  G.     1964.     Sediments  as  Substrates.     In  Imbrie,  J.  &  Newell,  J.     Approaches  to 

Palaeoecology  :  238-271,  16  figs.     New  York. 
Raciborski,  M.     1890.     Taonnrus  ultimus  Sap.  in  Galizien.    Verh.  k.  k.  geol.  Reichsanst.,  Wien, 

1890  :  265-266. 
Reid,  C.     1898.     The  Geology  of  the  Country  around  Eastbourne.     Mem.  Geol.  Surv.  U.K., 

London,    iv  +  15  pp.,  18  figs. 
Reis,   O.   M.     1910.     Beobachtungen  iiber  Schichtenfolge  und   Gesteinsausbildungen  in   der 

frankischen  Trias.     I.     Muschelkalk  und  untere  Lettenkohle.     Geognost.  Jahresh.,  Miinchen, 

22  :  1-258,  pis.  i-ii. 
1922.     Uber  Bohrrohren  in  fossilen  Schalen  und  iiber  Spongeliomorpha.     Z.  deutsch.  geol. 

Ges.,  Berlin,  73  :  227-237,  pi.  7. 
Richter,   P.  B.     1909.     Beitrdge  zur  Flora  der  unteren  Kreide  Quedlinburgs .     Teil  II.     Die 

Gattung  Nathorstiana  P.  Richter  und  Cylindrites  spongioides  Goeppert.     12  pp.,  pis.  8-12. 

Leipzig. 
Richter,  R.     1947.     In  O.  Wilkens,  Palaontologische  und  geologische  Ergebnisse  der  Reise 

von  Kohl-Larsen  (1928-29)  nach  Sud  Georgien.    Abh.  senckenberg.  naturf.  Ges.,  Frankfurt, 

474  :  75,  9  pis. 
Rieth,  A.     1932.     Neue  Funde  Spongeliomorpher  Fucoiden  aus  dem  Jura  Schwabens.     Geol.- 

Palaont.  Abh.,  Jena,  19  :  257-294,  pis.  13-17. 
Rios,  J.  M.  &  Hancock,  J.  M.     1961.     Summer  Field  Meeting  in  the  Spanish  Pyrenees.     Proc. 

Geol.  Ass.,  London,  72  :  373-390,  pis.  15,  16. 
Rose,  C.  B.      1835-6.      A  sketch  of  the  Geology  of  West  Norfolk.     Phil.  Mag.,  London  (3) 

7  :  171-182,  274-279,  370-376;   8  :  28-42. 
Saporta,  G.  de     1882.     A  propos  des  algues  fossiles.     82  pp.,  9  pis.     Paris. 

1884.     Les  organismes  problematiques  des  anciennes  mers.     100  pp.,  13  pis.     Paris. 

1887.     Nouveaux  documents  relatifs  aux  organismes  proble-matiques  des  anciennes  mers. 

Bull.  Soc.  Geol.  Fr.,  Paris  (3)  15  :  286-302,  pis.  3-7. 
Schaefer,   W.     1962.     Aktuo-Palaontologie  nach  Studien  in  der  Nordsee.     666   pp.,    36  pis. 

Frankfurt  am  Main. 
Schlotheim,  E.  F.  Baron  v.     1820.     Die  Petrefactenkunde  auf  ihrem  jetzigen  Standpunkte  durch 

die  Beschreibung  seiner  Sammlung  versteinerter  und  fossiler  Vberreste  des  Thier-und  Pflanzen- 

reiches  der  Vorwelt  erldutert.     437  pp.,  15  pis.     Gotha. 
Schroeder,  H.  &  Boehm,  J.,  1909.     Geologie  und  Palaontologie  der  Subhercynen  Kreidemulde. 

Abh.  preuss.  geol.  Landesanst.,  Berlin  (n.f.)  56  :  1-64,  pis.  1-16. 
Schultze,  C.  F.     1760.     Betrachtung  der  versteinerten  Seesterne  und  ihrer  Theile.     iv  +  58  pp., 

3  pis.     Warsaw  &  Dresden. 
Seeley,  H.     1864.     On  the  Hunstanton  Red  Rock.     Quart.  Jl.  geol.  Soc.  Lond.,  20  :  327-332. 
Seilacher,  A.     1 95 1.     Der  Rohrenbau  von  Lanice  conchilega  (Polychaeta) .     Ein  Beitrag  zur 

Deutung  fossiler  Lebensspuren.     Senckenbergiana,  Frankfurt  am  Main.,  32  :  267-280. 
1955-     Spuren  und  Lebenweise  der  Trilobiten:    Spuren  und  Fazies  im  Unterkambrium. 

In  Schindewolf,  O.  H.  &  Seilacher,  A.,  Beitrage  zur  Kenntnis  des  Kambriums  in  der  Salt 

Range  (Pakistan).     Akad.  Wiss.  Lit.  Mainz,  math.^nat.  Kl.,  10  :  86-143,  pis.  16-27. 
1964.     Biogenic  sedimentary  structures.     In  Imbrie,    J.    &   Newell,   D.     Approaches  to 

Palaeoecology  :  296-316,  8  figs.,  1  tab.     New  York. 
Semenow,  W.  P.     1899.     La  Faune  des  d6p6ts  cre-taces  de  Mangychlak  et  de  quelques  autres 

localites  de  la  province  transcaspienne.     Trav.  Soc.  Imp.  Nat.  St.  Petersb.,  28,  5  (Geol.,  Min.)  : 

1-178,  pis.  1-4. 


LOWER  CHALK  TRACE  FOSSILS  OF  S.  ENGLAND  167 

Simpson,  S.     1957.     On  the  trace-fossil  Chondrites.     Quart.  Jl.  geol.  Soc.  Lond.,  112  :  475-499, 

pis.  21-24. 
Sorby,  H.  C.     1861.     On  the  organic  origin  of  the  so-called  "  Crystalloids  "  of  the  Chalk. 

Ann.  Mag.  Nat.  Hist.,  London  (3)  8,  193-200. 
Sowerby,  J.  de  C.     1823-25.     The  Mineralogy  Conchology  of  Great  Britain;   or  coloured  figures 

of  those  remains  of  Testaceous  Animals  or  Shells,  which  have  been  preserved  at  various  times, 

in  the  earth,  5,  pis.  408-443  (1823),  444-485  (1824),  486-508  (1825).     London. 
Sternberg,  K.  M.  Graf.  v.     1833.      Versuch  einer  geognostisch-botanischen  Darstellung  dcr  Flora 

der    Vorwelt.     5-6.     80   pp.     Leipzig   &   Prague. 
Stevens,  B.  A.     1928.     Ecological  observations  on  the  Callianassidae  of  Puget  Sound.    Ecology, 

New  York,  10  :  399-405. 
Taylor,  J.  H.     1964.     Some  aspects  of  diagenesis.     Advmt.  Sci.,  Lond.,  18  :  417-436,  7  figs. 
Taylor,   R.     1823.     Geological  section  at  Hunstanton  Cliff,   Norfolk.     Phil.  Mag.,  London, 

61  :  81-83. 
Thomel,  G.     1961.     Sur  la  presence  du  genre  Schloenbachia  dans  le  Vraconien  superieur  de 

Saint-Laurent  de  1'  Escarene    (Alpes-Maritimes).     C.R.  Soc.  Geol.  Fr.,  Paris,  1961  :  37. 
Verwey,    J.     1930.     Einiges    fiber    die    Biologie    ost-indischer    Mangrovekrabben.     Treubia, 

Buitenzorg,  Bogor,  12  :  167-261,  pis.  6-14. 
Watelet,  A.     1866.     Description  des  plantes  fossiles  du  basin  de  Paris,     iv  +  264  pp.,  Atlas 

30  pp.,  60  pis.     Paris. 
Webster,  T.     1814.     On  some  new  varieties  of  fossil  Alcyonia.      Trans,  geol.  Soc,  Lond.  :  371— 

387,  pis.  27-29. 
Weigelt,  J.     1929.     Fossile  Grabschachte  brachyurer  Decapoden  als  Lokalgeschiebe  in  Pom- 
mem  und  das  Rkizocorallium-Problem.     Z.  Geschiebforsch.,  Berlin,  5  :  1-42,  pis.  1-4. 
Weimer,  R.  J.  &  Hoyt,  J.  H.     1964.     Burrows  of  Callianassa  major  Say,  geologic  indicators 

of  littoral  and  shallow  neritic  environments.     J.  Paleont.,  Tulsa,  38  :  761-767,  pis.  123,  124. 
Whitaker,  W.  &  Jukes-Browne,  A.  J.     1899.     The  Geology  of  the  Borders  of  the  Wash: 

including  Boston  and  Hunstanton.     Mem.  Geol.  Surv.  U.K.,  London,     iv  +  146  pp. 
White,  E.  I.     1923.     Notes  on  a  new  Species  of  Terebelloid  and  other  Phenomena  in  the  Great 

Pit  at  Harefield,  Middlesex,  etc.     Proc.  Geol.  Ass.,  London,  34  :  43-46. 
Wilcox,  N.  R.     1953.     Some  coprolites  from  phosphatic  chalks  in  S.E.  England.     Ann.  Mag. 

Nat.  Hist.,  London  (12)  6  :  369-375,  pi.  11. 
Wilson,  V.     1949.     The  Lower  Corallian  Rocks  of  the  Yorkshire  Coast  and  Hackness  Hills. 

Proc.  Geol.  Ass.,  London,  60  :  235-271,  pis.  9,  10. 
Wiltshire,  T.     1859.     On  the  Red  Chalk  of  England.     Geologist,  London,  2  :   261-278   and 

Geol.  Ass.  Spec.  Papers,  1859  :   1-18,  pis.  1-4. 

1869.     On  the  Red  Chalk  of  Hunstanton.     Quart.  Jl.  geol.  Soc.  Lond.,  25  :  185-192. 

Wood,  C.  J.     1965.     Field  Meeting  on  the  Lower  Chalk  between  Folkestone  and  Dover.     Proc. 

Geol.  Ass.,  London,  76  :  301-304. 
Woodward,  S.     1830      A  synoptic  table  of  British  Organic  remains,     xiii  +  50  pp.,  1  pi.     Lon- 
don &  Norwich. 

1833.     An  Outline  of  The  Geology  of  Norfolk.     54  +  1  pp.,  6  pis.     Norwich. 

Zareczny,  S.  W.      1878.     O  srednich  warstwach  kredowych  w  krakowskim  okregu.     Krakow 

Kom.  Fizyogr.  Spraw.,  12  :  176-246,  pis.  4-8. 


PLATE  i 

Fig.  i.  Thalassinoides  saxonicus  (Geinitz).  Lower  Chalk,  Middle  Cenomanian;  Houghton 
Regis,  near  Dunstable,  Beds.  Base  of  a  fallen  block  of  Totternhoe  Stone  with  part  of  the  under- 
lying Chalk  Marl  attached,  the  burrow  is  filled  by  Totternhoe  Stone.  Plan  view,  hammer  head 
16  cm.  long. 

Fig.  2.  Thalassinoides  cf.  suevicus  (Rieth).  Upper  Greensand;  Cow  Gap,  N.E.  of 
Beachy  Head,  Eastbourne,  Sussex.  Burrow  originates  from  the  base  of  the  Glauconitic  Marl 
(Lower  Cenomanian).     Plan  view,  scale  in  inches. 

Fig.  3.  Burrows  at  base  of  the  Glauconitic  Marl,  Lower  Cenomanian;  Compton  Bay,  Isle  of 
Wight.     Vertical  section,  hammer  head  16  cm.  long. 

Fig.  4.  Laminated  structures.  Lower  Chalk,  Middle  Cenomanian,  bed  7 ;  foot  of  cliff  600  m. 
E.  of  Akers  steps,  Dover,  Kent.     Vertical  section,  pencil  9  cm.  long. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE   1 


■I 


X 


GEOL.  15,  3. 


17 


PLATE  2 

Fig.  i.  Spongeliotnorpha?  annulatum  ichnosp.  nov.  Glauconitic  Marl,  Lower  Cenoma- 
nian;   foreshore,  East  Wear  Bay,  Folkestone,  Kent.  Oblique  section,  hammer  head  16  cm.  long. 

Fig.  2.  Vertical  section  of  burrowed  chalk  with  abundant  Chondrites  sp.  Lower  Chalk, 
Upper  Cenomanian;    Betchworth  Limeworks,  Betchworth,  Surrey,      x  i. 

Fig.  3.  Vertical  section  of  burrowed  chalk  with  Chondrites  sp.  and  burrow  type  E.  Lower 
Chalk,  Middle  Cenomanian;    Eastbourne,  Sussex,      xi. 

Fig.  4.  Vertical  section  of  burrowed  chalk.  Lower  Chalk,  Upper  Cenomanian;  Dover, 
Kent.     Note  relative  abundance  of  Chondrites  sp.  in  larger  burrows,      xi. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE  2 


geol.  15,  3. 


i7§ 


PLATE  3 

Thalassinoid.es  paradoxica  (Woodward) 

Neotype,  B.M.  (N.H.)  T.545.     Lower  Chalk,  Lower  Cenomanian,  Paradoxica  bed;    Hunstan- 
ton, Norfolk.     Oblique  view,  scale  in  centimetres. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE  3 


%'■■■ 

i 

PLATE  4 

Thalassinoides  paradoxica  (Woodward) 

Neotype,  B.M.  (N.H.)  T.545.     Lower  Chalk,  Lower  Cenomanian,  Paradoxica  bed;    Hunstan- 
ton, Norfolk.     Plan  view,  scale  in  centimetres. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE  4 


1  -,*>  P*^  .,  ■■ 


PLATE  5 

Fig.  i.  Laminated  structure  showing  disruption  of  lamination  by  subsequent  burrowing. 
B.M.  (N.H.)  T.550,  Lower  Chalk,  Middle  Cenomanian;  Pit  N.E.  of  Wouldham  Hall,  Wouldham, 
Kent.     Vertical  section,   x  1.     (Detail  of  PI.  8,  fig.  3). 

Fig.  2.  Thalassinoides  saxonicus  (Geinitz).  B.M. (N.H.)  T.547,  Lower  Chalk,  Middle 
Cenomanian,  chalk  beneath  Totternhoe  Stone;  Houghton  Regis,  near  Dunstable,  Beds.  Plan 
view  of  upper  surface  of  termination  showing  ornamentation  of  elongate  mounds.      X  f . 

Fig.  3.  Thalassinoides  saxonicus  (Geinitz).  B.M. (N.H.)  T.548.  Same  horizon  and 
locality,  detail  of  figured  specimen  (PI.  6,  fig.  3),  showing  bottom  covered  with  Chondrites  s$.     Xi. 

Fig.  4.  Burrow  type  D.  S.M.C.  bg2473  (Jefferies  collection).  Plenus  Marls,  bed  i;  Merst- 
ham,  Surrey,      x  1. 

Fig.  5.  Spongeliomorpha?  annulatum  ichnosp.  nov.  Holotype,  B.M.  (N.H.)  T.554. 
Glauconitic  Marl,  Lower  Cenomanian;  section  below  Martello  Tower  no.  3,  Copt  Point, 
Folkestone,  Kent.      Xi. 

All  specimens  except  Fig.  1  coated  with  ammonium  chloride. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE  5 


y_ 


5r> 


MS? 

;  .  -£f    St  I    vj£ 


I .;/ 


/j 


!r»  a 


&&*3&j&A 


r&* 


m^iMmn  mm 


u  m 


m 
iA 

■M0&  .'>'« 


Nk 


A'     if 


m 


m 


PLATE  6 

Fig.  i.  Pseudobilobites  jefferiesi  ichnosp.  nov.  B.M.  (N.H.)  T.556.  Lower  Chalk,  Upper 
Cenomanian;  7  m.  below  top  of  bed  8,  base  of  Shakespeare  Cliff;  W.  of  Dover,  Kent.  Bottom 
surface,   XI. 

Fig.  2.  Cylindrical  burrow  full  of  coarse  debris.  S.M.C.  B92827  (J efferies  collection).  Plenus 
Marls,  bed  i;    Merstham,  Surrey.     Listed  by  Jefferies  (1963)  as  Problematicum  sp.      xi. 

Fig.  3.  Thalassinoides  saxonicus  (Geinitz).  B.M. (N.H.)  T.548.  Lower  Chalk,  Middle 
Cenomanian,  chalk  beneath  Totternhoe  Stone;  Houghton  Regis,  near  Dunstable,  Beds.  Bottom 
surface  of  typical  branching  fragment  covered  with  Chondrites  sp.      x  1. 

Fig.  4.  Thalassinoides  ornatus  ichnosp.  nov.  associated  with  T.  saxonicus  (Geinitz). 
Holotype  B.M. (N.H.)  T.559.     Same  horizon  and  locality.     Top  surface,    x  1. 

All  specimens  coated  in  ammonium  chloride. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE  6 


Wf 


PLATE  7 

Fig.  i.  Burrow  type  D.  B.M.  (N.H.)  T.557.  Lower  Chalk,  Middle  Cenomanian,  horizon 
of  abundant  Orbirhynchia  mantelliana  (Sowerby)  and  Sciponoceras  baculoide  (Mantell) ; 
Beddingham  Limeworks,  Beddingham,  near  Glynde,  Sussex,      x  1. 

Fig.  2.  "  Terebella "  harefieldensis  White.  Holotype,  B.M.(N.H.)  A.2445.  Chalk/ 
Reading  Beds  junction;    The  Great  Pit,  Harefield,  Middlesex,      x  1. 

Fig.  3.  Pseudobilobites  jefferiesi  ichnosp.  nov.  S.M.C.  B91035.  Plenus  Marls,  bed  i; 
Merstham,  Surrey.     Figured  Jefferies  (1963,  pi.  77,  fig.  5).     Bottom  surface,    xi. 

Fig.  4.  Burrow  type  D.  B.M. (N.H.)  T.569,  Lower  Chalk,  Middle  Cenomanian;  Bluebell 
Hill,  Burham,  Kent,  x  1 . 

Fig.  5.  Burrow  type  D.  B.M. (N.H.)  T.558,  Lower  Chalk,  Middle  Cenomanian;  Glynde, 
Sussex,      xi. 

Fig.  6.  Thalassinoides  ornatus  ichnosp.  nov.  Paratype,  B.M.  (N.H.)  T.551.  Lower 
Chalk,  Middle  Cenomanian,  chalk  below  Totternhoe  Stone;  Houghton  Regis,  near  Dunstable, 
Beds,      x  1. 

Fig.  7.     Spongeliomorpha  sp.     B.M.  (N.H.)  T.553,  same  horizon  and  locality,      x  1. 

All  specimens  except  Fig.  2  coated  in  ammonium  chloride. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


.1 


* 


','< 


w 


&ss 


PLATE  8 

Fig.  i.  "  Terebella  "  cancellata  Bather.  Holotype,  B.M.  (N.H.)  58253  (Capron  collection). 
Lower  Chalk,  subglobosus  Zone;    Glynde,  Sussex.     Figured  Bather  (1911,  pi.  24,  fig.  3).      xi. 

Fig.  2.  "  Terebella"  cancellataBather.  Paratype,  B.M.  (N.H.)  A. 1574  (Capron collection). 
Lower  Chalk;    Cowslip  pit,  near  Guildford,  Surrey.      X  1. 

Fig.  3.  Laminated  structure.  B.M.  (N.H.)  T.550,  Lower  Chalk,  Middle  Cenomanian;  pit 
N.E.  of  Wouldham  Hall,  Wouldham,  Kent.  Vertical  section,  upper  surface  at  left  margin. 
Xf. 

Fig.  4.  Pseudobilobites  jefferiesi  ichnosp.  nov.  S.M.C.  B91557  (Jefferies  collection). 
Plenns  Marls,   bed  i;    Merstham,   Surrey.     Lower  surface,  X 1. 

Fig.  5.  Thalassinoides  paradoxica  (Woodward)  B.M.  (N.H.)  T. 549.  Lower  Chalk,  Lower 
Cenomanian,  Paradoxica  bed ;  Hunstanton,  Norfolk. 

Figures  1-3  uncoated,  4,  5  coated  with  ammonium  chloride. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  3 


PLATE  8 


PLATE  9 

Fig.  i.  Chondrites  sp.  B.M.  (TST.H.)  T.562.  Lower  Chalk,  Middle  Cenomanian,  horizon 
of  abundant  Orbirhynchia  mantel  liana  (Sowerby)  and  Sciponoceras  baculoide  (Mantell) ;  300  m. 
west  of  Head  Ledge,  N.E.  of  Beachy  Head,  Eastbourne,  Sussex.      x  1. 

Fig.  2.  Thalassinoides  paradoxica  (Woodward)  B.M.  (N.H.)  T.546,  Lower  Chalk,  Lower 
Cenomanian,  Paradoxica  bed;   Hunstanton,  Norfolk,      x  1. 

Figs.  3,  4.  Pseudobilobites  jefferiesi  ichnosp.  nov.  Holotype  B.M.  (N.H.).  T.565.  Lower 
Chalk,  Middle  Cenomanian,  chalk  below  Totternhoe  Stone;  Pitstone,  Buckinghamshire.  3, 
upper  surface,  4,  lower  surface,      x  1. 

Fig.  5.  Burrow  type  D.  S.M.C.  B92472  (Jefferies  collection).  Planus  Marls,  bed  i;  Merst- 
ham,  Surrey. 

Fig.  6.  Pseudobilobites  jefferiesi  ichnosp.  nov.  S.M.C.  Bgi653b  (Jefferies  collection). 
Plenus  Marls,  bed  1 ;  Lockinge,   Berkshire. 

All  specimens  except  Figs.  1  and  6  coated  with  ammonium  chloride. 


Bull.  Br.  Mus.  nat.Hist.  (Geol.)  15,  3 


PLATE  9 


■ 

rP^'. 

H  ■  1 

* 

m 

it - 

V 

St 

f:- 

J 

'  Jr 

2 


PRINTED  IN  GREAT  BRITAIN 
BY  ADLARD  &  SON  LIMITED 
BARTHOLOMEW    PRESS,    DORKING 


A  NEW   TEMPSKYA  FROM  KENT 


M.  E.  J.  CHANDLER 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol  15  No.  4 

LONDON:   1968 


A  NEW   TEMPSKYA  FROM  KENT 


BY 

MARJORIE  E.  J.  CHANDLER 


P/>.  169-179;  12  P/ates 


BULLETIN  OF 

THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 

GEOLOGY  Vol.  15  No.  4 

LONDON:  1968 


THE    BULLETIN    OF    THE    BRITISH    MUSEUM 

(natural  history),  instituted  in  1949,  is 
issued  in  five  series  corresponding  to  the  Departments 
of  the  Museum,  and  an  Historical  series. 

Parts  will  appear  at  irregular  intervals  as  they  become 
ready.  Volumes  will  contain  about  three  or  four 
hundred  pages,  and  will  not  necessarily  be  completed 
within    one    calendar  year. 

In  1965  a  separate  supplementary  series  of  longer 
papers  was  instituted,  numbered  serially  for  each 
Department. 

This  paper  is  Vol.  15,  No.  4  of  the  Geological 
Palaeontological  series.  The  abbreviated  titles  of  the 
periodicals  cited  follow  those  of  the  World  List  of 
Scientific  Periodicals. 


World  List  abbreviation 
Bull.  Br.  Mus.  nat.  Hist.  (GeoL). 


©  Trustees  of  the  British  Museum  (Natural  History)  1968 


TRUSTEES    OF 
THE    BRITISH    MUSEUM     (NATURAL    HISTORY) 

Issued  12  January,  1968  Price  £1  10s 


A  NEW   TEMPSKYA  FROM  KENT 

By  M.  E.  J.  CHANDLER 

MS  accepted  18th  April,  1967 

SYNOPSIS 

A  silicified  specimen  with  well-preserved  cell  structure,  found  on  the  shore  at  Sheppey,  proves 
to  be  a  new  species  of  the  Mesozoic  "  genus  "  Tempskya  in  which  individual  solenostelic  stems  are 
welded  together  into  a  "  false  stem  "  by  their  intertwining  roots.  Detailed  morphology  links 
this  specimen  most  closely  with  Tempskya  grandis  from  the  Upper  Cretaceous  of  Wyoming  but 
the  two  are  distinguished  by  the  greater  number  of  meristeles  normally  present  within  the  rhi- 
zome section  of  the  American  species  and  the  great  difference  of  size. 

The  type  of  siliceous  preservation  suggests  that  the  specimen  may  have  come  from  the 
Woolwich  Beds  of  Heme  Bay. 

The  Kent  Tempskya  is  only  the  second  species  to  be  recorded  from  England  and  is  quite 
distinct  from  the  well  known  Tempskya  erosa  found  in  the  Wealden  Beds  and  the  Lower  Green- 
sand. 

INTRODUCTION 

Some  years  ago  a  member  of  the  United  States  Geological  Survey,  Dr.  R.  A.  Scott, 
paid  a  visit  to  the  Sheppey  coast  to  see  this  famous  source  for  London  Clay  plants. 
He  picked  up,  lying  loose  on  the  shore,  a  somewhat  waterworn  "  stem  "  and  kindly 
gave  it  to  the  British  Museum  (Natural  History).  A  transverse  cut  just  below  the 
apex  of  the  specimen  (V.51841)  showed  that  it  was  a  fern  beautifully  preserved  in 
silica  with  excellent  cell  structure.  The  siliceous  preservation  raised  the  question 
of  the  origin  of  this  plant  so  unlike  the  London  Clay  fruits  and  seeds  which  are 
commonly  pyritized,  occasionally  carbonaceous  or  with  calcific  internal  casts  and 
not  infrequently  a  mixture  of  pyrites  and  carbonaceous  tissues.  The  specimen  has 
now  been  studied  and  the  results  of  the  investigations  are  described  below. 

DESCRIPTION  OF  SPECIMEN 

The  silicified  "  false  stem  "  was  originally  about  12-5-13  cm.  long.  The  cutting 
process  when  the  apex  was  severed  of  course  involved  some  loss  of  length.  Both 
parts  of  the  specimen  have  now  been  examined,  further  cuts  have  been  made,  and  a 
few  thin  sections  have  been  prepared.  The  specimen  is  now  in  four  fragments.  The 
poorly  preserved  basal  part  is  numbered  V.51841,  the  central  part  V. 51841a; 
and  the  apex,  which  has  further  been  cut  longitudinally  into  two  fragments, 
V.51841&  and  c.  Of  the  six  slides  (V. 51841^-2'),  ¥.51841^  from  the  top  of  V.51841 
is  too  ill-preserved  to  be  very  informative  or  worth  thinning.  V.518410  and  / 
are  from  the  upper  surface  of  V. 51841a,  the  slide  e  being  lower  than/.  Slides 
V .  51841  g,  h  and  i  come  from  the  base  of  V .  51841&.  They  do  not  extend  across  the 
whole  breadth  of  the  specimen  nor  even  of  the  cut  surface  of  V.51841&,  but  all  three 
together  with  b  1  and  2  lie  within  about  10  mm.  of  the  length  of  the  "  stem  ". 

GEOL.   15,  4.  l8§ 


1 72  A   NEW   TEMPSKYA    FROM   KENT 

The  transverse  diameter  of  the  whole  specimen  at  broadest  (i.e.  at  about  the  middle) 
is  5.4  by  3.5  cm.  In  its  rolled  and  waterworn  state  it  shows  superficially  a  complex 
of  stems  with  general  longitudinal  alignment  and  some  evidence  of  dichotomous 
bifurcation  and  intertwining,  (PL  1,  figs.  1,  2).  Remains  of  a  few  projecting 
scattered  petiole  bases  are  seen  (PI.  1,  fig.  2;  PL  2,  fig.  4).  Section  V. 51841/ passes 
through  one  stem  in  the  early  stages  of  dichotomous  division  (PL  1,  fig.  3;  PL  3, 
fig.  6;  PL  4,  fig.  7)  so  that  there  are  two  steles  but  both  are  still  surrounded  by  a 
single  cortex  and  epidermis  which  have  become  somewhat  bilobed.  The  whole 
specimen  is  more  abraded  on  one  broad  surface  than  on  the  other.  Some  crushing, 
dislocation  and  disintegration  can  be  detected  in  the  section  towards  this  worn  side 
making  the  structures  more  obscure  here  than  elsewhere.  There  has  been  some 
disruption  of  the  stele,  for  example,  in  the  dividing  stem  (PL  3,  fig.  6;  PL  4,  fig.  7). 
A  few  deep  concavities  show  the  surface  view  of  ramentae  in  lighter  coloured  silica 
(PL  1,  fig.  1).  They  are  obscured  by  abundant  freely  branching  roots  many  of  which 
grew  upwards.  The  fringed  edges  of  the  scales,  due  to  their  multi-cellular  structure, 
show  clearly  in  places.  Despite  wear  and  tear  the  specimen  appears  to  retain, 
approximately  at  least,  the  original  length  and  breadth,  for  emerging  petiole  bases 
project  from  the  general  surface  both  on  the  side  and  at  the  apex.  The  structure, 
at  the  lower  end  where  amorphous  silica  has  obliterated  cell  tissues,  is  more  obscure 
than  in  the  upper  two- thirds  of  the  specimen.  More  especially  is  this  the  case  on  the 
less  well  preserved  side. 

In  transverse  section  in  the  upper  part  (PL  1,  fig.  3)  all  or  part  of  seven  radially 
arranged  dorsiventral  stems  are  visible.  They  show  a  typical  solenostelic  structure. 
Five  of  them  are  closely  adjacent  to  or  actually  at  the  circumference;  two  are  more 
deeply  embedded  but  still  radially  aligned.  In  all,  the  steles  are  towards  the  inner 
end  of  a  radial  line,  and  the  leaf  traces  or  meristeles  at  their  outer  ends.  The  spaces 
between  the  stems  are  packed  closely  with  ramentae  and  roots  of  various  sizes.  The 
latter  may  be  sectioned  longitudinally  for  a  short  distance  but  are  commonly 
transversely  or  obliquely  cut  in  the  slides  and  cut  surfaces.  The  penetrating  roots 
which  pierce  the  scales  and  weave  in  and  out  weld  the  stems  into  the  "  false  stem  ". 
The  solidity  of  the  whole  complex  is  undoubtedly  enhanced  by  the  process  of  silicific- 
ation  which  unites  stems,  scales  and  roots  into  a  solid  inseparable  mass,  silica 
penetrating  all  tissues  and  replacing  cell  contents  just  as  in  Osmunda  dowkeri. 

The  maximum  transverse  diameter  of  the  component  stems  is  about  15.5  by 
9  mm.,  or  15  by  11  mm.  in  the  middle  of  the  specimen,  larger,  some  22  to  25  by 
11  mm.,  in  the  basal  part.  Near  the  apex  some  stems  are  smaller,  12  or  13  by  8  mm., 
but  others  are  15  by  13  and  16  by  11  mm. 

In  addition  to  the  roots  which  belong  to  the  fern  itself,  numerous  "  foreign  "  roots 
are  visible  in  the  sections.  In  particular  they  are  concentrated  in  an  irregular  but 
continuous  thin  belt  of  sclerenchyma  outside,  concentric  with  and  close  to  the  steles 
and  meristeles.  They  make  this  belt  of  tissue  very  conspicuous  while  often  obscuring 
its  detailed  structure.  They  are  also  abundant  in  the  sclerenchyma  of  the  pith 
adjacent  to  the  leaf  gaps  (PL  5,  fig.  8;  PL  7,  fig.  10;  PL  8,  fig.  11;  PL  10,  fig.  13). 
Some  of  these  roots  may  be  monocotyledonous,  others  dicotyledonous,  as  they  show 
radially  arranged  small  xylem  cells  containing  large  scattered  vessels.     I  am  indebted 


A    NEW    TEMPSKYA    FROM    KENT  173 

to  Dr.  Holttum  for  these  suggestions.  The  roots  sometimes  take  the  place  of  the 
stelar  tissues  of  the  fern  roots  within  a  dense  belt  of  surrounding  sclerenchyma. 

Structure  of  individual  stems.  The  stems  are  clearly  solenostelic.  Some 
sections  show  a  complete  cylinder  with  no  leaf  gap  (PI.  8,  fig.  n),  others  show  a 
single  leaf  gap  (PI.  7,  fig.  10;  PL  9,  fig.  12).  Yet  others  have  two  leaf  gaps  (PL  5, 
fig.  8 ;  PL  6,  fig.  9).  The  pattern  due  to  the  development  of  the  leaves  is  repeated  at 
regular  intervals  in  the  individual  stems  so  that  the  sections  made  at  different  levels 
across  the  false  "  stem  "  show  successive  changes.  In  any  single  section  the  in- 
cluded stems  are  commonly  at  different  stages  in  development.  Intermediate  stages 
have  been  seen  on  the  polished  surfaces  exposed  both  before  and  after  section 
cutting  but  being  unsuitable  for  photography  they  are  not  represented  in  the 
plates. 

The  periphery  of  the  stems,  inside  an  ill-preserved  epidermis  from  which  the 
ramentae  arise,  is  formed  of  several  layers  of  opaque  sclerenchymatous  cells  with  dark 
contents.  Up  to  a  dozen  layers  have  been  counted.  Exactly  similar  outer  tissues 
are  described  by  Read  &  Brown  (1937  :  no)  in  Tempskya  grandis  as  the  "  outer 
cortex  ".  Within  is  a  layer  of  sclerenchyma,  at  least  twenty  cells  thick,  cells  which 
although  thick-walled  retain  a  considerable  lumen.  It  clearly  corresponds  with  the 
"  middle  or  sclerenchymatous  cortex  "  in  T.  grandis.  Individual  cells  are  isodia- 
metric  in  cross  section  where  the  walls  appear  to  be  unevenly  patchily  thickened. 
Inside  again  occurs  a  thick  layer,  twenty  or  more  cells  thick,  of  thin-walled  paren- 
chyma (corresponding  to  the  "  inner  cortex  "  of  T.  grandis).  It  is  at  least  as  broad 
as  and  may  be  broader  than  the  middle  sclerenchymatous  cortex.  Almost  invariably 
the  individual  cells  show  black  rounded  objects  occupying  much  of  the  cavity; 
sometimes  these  objects  themselves  lie  in  a  rounded  cavity  within  the  cell.  Under 
the  quarter-inch  lens  most  of  them  appear  to  be  agglomerated  crystals,  although  a 
few  simple  crystals  are  seen.  In  longitudinal  sections  of  the  cells  they  appear  as 
clumps  of  elongate  crystals  with  their  long  axes  at  right  angles  to  the  longer  axes 
of  the  cells.  Similar  inclusions  occur  sparsely  in  the  sclerenchyma  of  the  middle 
cortex.  Towards  the  inner  margin  of  the  parenchymatous  zone  there  is  always  a 
continuous  narrow  band  of  sclerenchyma  very  irregular  in  thickness  and  therefore 
in  outline  as  seen  in  transverse  section.  It  may  be  sharply  delimited  from  the 
parenchyma  in  which  it  lies  by  a  dark  line  on  the  outer  side,  (PL  5,  fig.  8;  PL  8, 
fig.  n;  PL  9,  fig.  12),  but  less  sharply  from  three  or  four  layers  of  parenchyma  cells 
which  lie  between  it  and  the  stele.  These  innermost  parenchyma  cells  are  rather 
small  but  display  the  same  crystalline  inclusions.  They  also  occur  abundantly  in 
the  irregular  sclerenchyma  belt  just  described,  a  belt  seen  only  in  T.  grandis  among 
the  well  preserved  fossil  species  and  in  the  less  well  preserved  T.  superba  Arnold. 

The  stele  is  bounded  both  externally  and  internally  by  an  endodermis  and  associ- 
ated tissues.  In  places  the  endodermis  is  well  defined  as  a  single  layer  of  equiaxial 
cells.  Inside  it,  preserved  only  in  certain  places,  are  two  or  more  layers  of  tangenti- 
ally  elongate  thin-walled  cells  with  clear  cavities  (PL  11,  fig.  14).  They  probably 
represent  pericycle  and  phloem  but  no  sieve  plates  appear  to  be  preserved.  The 
xylem  varies  considerably  in  thickness  from  about  three  tracheids  in  depth  to  sixteen 
or  twenty  as  a  result  of  leaf  trace  formation.     Along  the  margins  of  the  xylem  are 


174  A   NEW    TEMPSKYA    FROM    KENT 

patches  of  very  small  tracheids,  some  certainly  showing  scalariform  thickening.  The 
bulk  of  the  xylem  is  composed  of  conspicuous  large  metaxylem  tracheids  with  multi- 
seriate  thickening  which  causes  them  to  appear  angular  in  transverse  section  so 
producing  a  characteristic  pattern.  Where  roots  are  about  to  arise  the  small  elem- 
ents at  the  margin  become  more  numerous  causing  an  outward  bulge  in  the  endo- 
dermis.  It  is  probable  that  some  of  these  are  protoxylem  but  no  spiral  thickening 
appears  to  be  preserved.  A  considerable  amount  of  parenchyma,  often  with 
crystalline  inclusions  is  scattered  among  the  tracheids,  again  as  in  T.  grandis. 
T.  rossica  and  T.  wesselii  also  show  this  feature  (Andrews  &  Kern  1947  :  147)  as 
does  T.  superba  (Arnold  1958  :  137).  Where  the  stele  becomes  thin  and  the  bulge 
which  initiates  a  leaf  trace  begins  to  form,  the  metaxylem  tracheids  follow  a  tangen- 
tial course  (PL  10,  fig.  13).  In  the  actual  slide  they  then  show  the  thickening  clearly. 
At  one  leaf  gap  the  end  of  the  stele  abutting  on  it  shows  what  appear  to  be  some 
spiral  tracheids  near  the  margin  of  the  xylem. 

Where  the  pith  within  the  stele  adjoins  the  endodermis  there  is  a  thin  layer  of 
parenchyma  with  the  usual  inclusions.  It  is  continuous  through  the  leaf  gap  with 
the  fine  parenchyma  which  bounds  the  stele  externally.  Otherwise  almost  the 
whole  pith  is  formed  of  sclerenchyma  with  thick-walled  cells,  and,  at  the  centre,  with 
little  lumen  (PI.  5,  fig.  8 ;  PI.  7,  fig.  10).  The  sclerenchyma  is  also  continuous  through 
the  gap  with  the  sclerenchyma  of  the  cortical  zone  of  the  rhizome. 

Scales  or  Ramentae.  These  may  be  very  broad.  One  which  could  be  measured  is 
at  least  10  mm.  wide.  The  elongate  cells  which  form  them  he  end  to  end  and  side  to 
side.  In  the  rows  of  cells  the  end  walls  are  transverse  or  oblique  to  the  length.  The 
cell  rows  diverge  towards  the  lateral  margins  of  the  scales  where  the  free  ends  of  the 
rows  separate  and  form  the  characteristic  fringed  edge.  Seen  in  transverse  section 
of  the  stems  they  appear  as  thin  multicellular  plates  of  tissue  (PL  2,  fig.  5;  PL  7, 
fig.  10). 

Development  of  leaf  traces  and  structure  of  meristele.  On  that  side  of  the  stele 
towards  the  circumference  of  the  "  false  stem  ",  leaf  gaps  arise  at  short  longitudinal 
intervals  indicating  that  the  leaves  must  have  been  crowded.  In  a  single  section 
across  one  stem  two  and  sometimes  three  leaf-traces  may  still  be  included  within  the 
cortex  and  epidermis  showing  that  they  arise  at  acute  angles  (PL  3,  fig.  6,  stem  b; 
PL  9,  fig.  12).  Once  they  have  emerged  from  the  stem  they  apparently  change  their 
direction  and  pass  out  of  the  "  false  stem  "  quickly.  An  occasional  projecting  leaf 
base  on  the  side  of  the  upright  "  false  stem  "  (PL  1,  fig.  2)  points  to  the  fact  that  they 
were  borne  at  intervals  along  its  length.  A  terminal  crown  may  also  have  occurred 
(PL  2,  fig.  4).  The  appearance  of  the  stele  in  transverse  section  varies  with  the 
degree  of  development  of  the  leaf  trace,  the  positions  of  the  leaf  gaps  and  meristeles 
in  successive  sections  suggesting  that  they  were  borne  in  two  somewhat  irregular 
rows.  The  stages  in  trace  formation  have  been  pieced  together  from  the  various 
stems  in  thin  sections  and  cut  surfaces.  Serial  sections  were  not  attempted  because 
there  is  only  one  specimen  and  its  preservation  is  so  patchy.  This  caused  the 
sections  to  break  irregularly  and  made  complete  transverse  slices  difficult  to  obtain. 
It  also  limited  the  amount  of  thinning  that  could  be  carried  out. 

Development  of  a  trace  is  heralded  first  by  the  thinning  of  part  of  the  stele  con- 


A    NEW   TEMPSKYA    FROM    KENT  175 

nected  with  its  formation  while  simultaneously  an  adjacent  protuberance  occurs 
(PL  6,  fig.  9,  rhizome  a;  PL  7,  fig.  10).  The  bulge  or  protuberance  then  develops 
a  pair  of  angular  thickenings  on  the  inside  of  the  stele  near  its  inner  limits  (PL  3, 
rig.  6,  rhizome  b).  Further  thinning  at  the  inner  ends  of  the  bulge  just  beyond 
the  thickened  angles  now  produces  a  gap,  first  on  one  side  and  then  on  the  other 
(PL  7,  fig.  10).  As  a  result  a  C-shaped  trace  opening  inwards  is  separated  from  the 
stele  (PL  9,  fig.  12;  cf.  PL  6,  fig.  9,  rhizome  a  with  PL  7,  fig.  10).  The  trace  passes 
upwards  and  outwards  through  the  cortex.  The  angles  on  the  inner  side  of  the  stele 
soon  become  elongated  tangentially  and  approach  one  another  fusing  as  they  pass 
up  the  stem  with  resulting  restoration  of  the  complete  cylinder. 

At  certain  stages  of  development  two  gaps  are  seen  in  a  single  transverse  section. 
In  this  case  a  semicircular  section  of  the  stele  with  its  flattened  surface  towards  the 
periphery  of  the  "  false  stem  "  gives  rise  to  thin  outward  curving  lobes  at  its  angles. 
The  lobes  separate  first  from  the  central  area  remaining  attached  at  their  inner  ends 
and  so  producing  a  double  convex  curve  on  each  side,  the  curve  of  the  developing 
trace  being  much  shorter  than  that  of  the  parent  stele.  The  central  fragment  of  stele 
which  may  be  simple  or  bilobed  then  appears  as  an  island  separated  from  the  main 
stele  by  two  developing  leaf  gaps  (PL  5,  fig.  8;  PL  6,  fig.  9,  stem  b).  Usually  one 
trace  becomes  detached  as  on  the  right  in  PL  6,  fig.  9,  while  the  other  still  remains 
attached  although  much  thinned  and  ready  for  almost  simultaneous  separation. 

Recently  emerged  leaf  traces  soon  form  elongate  bulges  on  the  stem  surface  even 
prior  to  their  complete  departure  from  it  (PL  5,  fig.  8,  cf.  m  in  PL  7,  fig.  10  and  in 
PL  9,  fig.  12).  The  final  separation  of  the  trace  as  it  passes  through  the  cortical 
tissues  next  occurs.  A  tongue  of  sclerenchyma  grows  inward  between  stem  and 
trace,  one  side  giving  off  a  branch  which  enters  the  bay  of  the  meristele.  Scleren- 
chyma from  the  opposite  side  also  extends  inwards  until  the  two  sides  meet  thereby 
completely  surrounding  both  stem  and  leaf  base  (m  in  PL  7,  fig.  10).  At  a  slightly 
higher  level  the  outermost  stem  tissues  grow  in  also  and  divide  sclerenchyma  of  stem 
and  leaf  base  so  that  they  completely  part  company  (p  in  PL  7,  fig.  10)  the  trace 
continuing  upwards  and  outwards  until  it  emerges  at  the  surface  of  the  "  false  stem  ". 

In  shape  the  meristele  becomes  markedly  incurved  as  it  passes  out  of  the  stem 
(contrast  the  three  meristeles  in  PL  9,  fig.  12)  while  simultaneously  its  ends  thicken. 
As  is  to  be  expected,  the  succession  of  tissues  is  as  in  the  stem  itself.  The  meristele 
may  be  only  one  or  two  tracheids  thick  at  the  middle  of  the  arc  but  there  are  up 
to  about  seven  at  the  incurved  ends.  Endodermis  and  associated  tissues  completely 
surround  it.  Owing  to  the  angle  at  which  traces  arise  transverse  sections  cut  them 
somewhat  obliquely  so  that  the  tissues  appear  blurred.  The  thin  parenchyma  which 
completely  encircles  the  stele  of  the  leaf  trace  and  its  irregular  sclerenchyma  is 
continuous  with  the  parenchyma  of  the  stem  until  separated  by  the  ingrowing  of  the 
sclerenchyma.  As  in  the  stem  the  cells  of  this  tissue  show  the  characteristic 
crystalline  inclusions. 

An  unexplained  peculiarity  is  seen  in  the  meristele  of  one  thin  section.  Some  of 
the  tissues,  more  especially  the  parenchyma  cells,  have  broken  down  and  in  their 
place  oval  opaque  black  bodies  can  be  seen.  Similar  structures  in  a  root  of  Tempskya 
knowltoni  from  Montana  are  described  and  figured  by  Seward  (1924  :  494,  pi.  17, 


176  A   NEW   TEMPSKYA    FROM   KENT 

fig.  24).  He  was  unable  to  explain  them  but  suggested  that  they  might  be  coprolites 
of  a  small  insect  or  possibly  escaped  cell  contents.  Certainly  in  the  Kent  specimen 
each  such  body  occupies  a  separate  cell  until  the  surrounding  cell  walls  have  actually 
broken  down.  Seward  adds  that  entomologists  he  consulted  were  unable  to  identify 
the  bodies  with  the  activities  of  any  known  boring  animal  and  no  trace  of  any  insect 
had  been  found.  "  They  consist  ",  he  stated,  "  of  finely  comminuted  plant  debris  or 
dark  masses  of  rounded  cell  contents  and  are  certainly  not  spores  ". 

The  roots  (PI.  2,  fig.  5).  These  vary  greatly  in  size  and  are  branched  repeatedly. 
Some  are  as  much  as  2  mm.  broad.  Many  roots  and  their  branches  grow  upward 
through  the  tissues.  They  show  a  small  stele  with  typical  diarch  arrangement,  a 
well-defined  endodermis  surrounding  it.  There  are  about  four  to  six  large  metaxylem 
tracheids,  flanked  at  opposite  poles  by  groups  of  about  three  to  six  small  protoxylem 
tracheids.  The  metaxylem  tracheids  may  be  0  •  027  mm.  or  less  in  diameter.  Phloem 
is  rarely  preserved.  Outside  the  endodermis  are  several  layers  of  concentrically 
arranged  sclerenchymatous  cells  with  well  developed  cavities.  They  are  succeeded 
further  out  by  a  thick  band  of  dense  sclerenchyma  with  cavities  obliterated.  In 
transverse  section  these  cells  appear  both  radially  and  concentrically  aligned  forming 
a  very  conspicuous  band  of  tissue  which  may  be  0-3  mm.  broad.  It  corresponds  to 
the  "  middle  cortex  "  of  Andrews  &  Kern  in  their  description  of  Tempskya  (1947  : 
139).  In  young  roots  the  sclerenchyma  may  be  less  dense.  Sometimes  outside  the 
sclerenchyma  there  is  another  concentric  belt  of  thin-walled  cells  of  about  the  same 
width  (Andrews'  "  outer  cortex  ").  It  is  not  invariably  preserved.  In  PI.  2,  fig.  5, 
the  "  outer  cortex  "  has  a  lozenge  or  diamond-shaped  outline  of  which  half  only  is 
preserved.  This  shape  is  dictated  by  the  pressure  of  closely  compacted  masses  of 
roots  in  the  spaces  between  the  stems.  In  some  roots  only  the  sclerotic  tissue 
survives  surrounding  an  empty  circular  space,  occasionally  occupied  by  foreign 
tissues  as  described  above. 

Affinities.  The  composite  character  of  this  "  false  stem  "  with  its  dichotomous 
solenostelic  true  stems  embedded  in  a  mass  of  their  own  roots  and  scales  connects 
the  specimen  with  the  Mesozoic  Tempskya  of  Corda  (1845  :  81).  Kidston  &  Gwynne- 
Vaughan  (191 1  :  13)  later  published  a  generic  diagnosis  of  the  genus  quoted  by 
Read  &  Brown  (1937  :  108).  Unfortunately  the  true  relationship  to  living  ferns 
has  not  yet  been  discovered  although  many  distinguished  botanists  have  carried  out 
research  on  the  subject.  As  long  ago  as  1872  Feistmantel  suggested  that  Tempskya 
was  not  a  genus  but  a  mode  of  preservation  of  several  distinct  types  of  fern  stems. 
To  the  writer,  this  view  appears  to  be  greatly  strengthened  by  the  occurrence  of  a 
supposed  Mesozoic  genus  in  the  Tertiary.  Read  &  Brown  (1937  :  120)  discuss  the 
taxomic  affinities  and  summarize  views  published  prior  to  their  paper.  They  created 
the  "  family  "  Tempskyaceae  as  the  natural  affinities  could  not  be  discovered  but 
probably  no  such  family  exists.  It  is  merely  a  convenient  way  of  grouping  different 
ferns  with  a  similar  habit.  Dr.  R.  E.  Holttum  and  Dr.  T.  G.  Walker  have  kindly 
examined  slides  or  detailed  photographs  of  the  Kent  fern  but  were  unable  to  recog- 
nize any  living  genus  with  which  there  is  complete  agreement.  It  is  necessary  to 
remember  that  the  appearance  of  these  plants  in  life  may  have  differed  materially 
from  that  when  fossilized  thanks  to  the  cementing  and  hardening  effects  of  silicifica- 


A   NEW   TEMPSKYA    FROM   KENT  177 

tion  welding  stems  and  roots  into  an  apparently  solid  entity.  Like  Feistmantel, 
Andrews  &  Kern  (1947  :  143)  consider  that  the  "  trunk  "  of  the  genus  "  represents  a 
peak  of  structural  evolution  that  is  manifest  in  a  generally  comparable  fashion  in  a 
number  of  ferns,  both  living  and  fossil  ".  They  refer  to  scattered  references  in 
literature,  not  quoted  in  detail,  to  living  ferns  having  an  upright  trunk  composed  of 
branching  stems  "  held  together  to  a  greater  or  lesser  degree  by  a  mass  of  adventiti- 
ous roots  ".  Among  these  are,  of  course,  some  of  the  well-known  tree  ferns  with  a 
similar  upright  trunk  or  caudex  which  differ  in  important  respects  from  Tempskya 
and  all  of  which  are  certainly  much  larger  than  the  Kent  specimen.  Writing  later 
Andrews  (1961  :  116)  after  briefly  describing  the  genus  commented  that  the  radial 
development  suggests  a  single  stem  in  the  early  sporeling  stage  which  grew  and 
divided  to  form  a  large  and  longer  trunk.  He  did  not  then  consider  the  anatomy  of 
the  stem  sufficiently  distinctive  to  afford  evidence  of  affinity  but  again  emphasized 
that  the  mode  of  growth  could  have  arisen  independently  in  several  unrelated 
groups.  None  known  to  him  agreed  closely  with  Tempskya  either  in  detailed  anat- 
omy or  in  the  huge  number  of  stems  involved.  The  smaller  size  of  the  Kent 
"  false  stem  "  with  its  fewer  true  stems  strongly  suggests  that  this  specimen  may 
eventually  be  matched  among  living  material  when  sufficient  knowledge  of  the 
anatomy  and  cytology  of  this  vast  group  of  plants  is  available  and  allowance  is  made 
for  the  profound  alteration  in  appearance  due  to  silicification.  Whatever  the 
relationship  of  the  Mesozoic  species  may  be  it  is  most  likely  that  the  Kent  specimen 
belongs  to  a  living  genus.  Naturally  Recent  root  stocks  and  stem  bases  are  not 
available  in  unlimited  amount  or  variety.  The  majority  of  Tempskya  species  are 
from  America  which  has  produced  at  least  seven  distinct  kinds.  Some  of  these  are 
enormous  attaining  to  16  in.  in  diameter  and  a  height  of  at  least  12  or  possibly 
19  or  20  feet.  Some  include  a  very  large  number  of  stems.  Thus  Andrews  &  Kern 
(1947  :  155)  quote  more  than  200  stems  in  a  single  trunk.  Some  of  the  American 
species  are  well  preserved  and  fall  into  two  clear  groups  (Read  &  Brown  1937  :  119; 
Andrews  &  Kern  1947  :  147).  The  Kent  "  stem  "  closely  agrees  in  its  detailed 
morphology  with  T.  grandis  Read  &  Brown  (1937  :  114,  pi.  29,  fig.  2;  pi.  32,  figs. 
2-5;  Pi-  33,  %s.  1-4;  pi.  34,  figs.  1-4;  pi.  35,  figs.  1-4;  pi.  41,  fig.  4;  pi.  42, 
figs.  1-3;  pi.  43,  figs.  1,  4-7)  which  belongs  to  the  first  of  these  groups.  It  is  an 
Upper  Cretaceous  species  from  the  Aspen  shales  of  Wyoming.  It  is  about  8  cm. 
in  diameter  and  some  20  cm.  long,  perhaps  larger  because  more  mature  than  the 
Kent  fossil.  Despite  its  somewhat  worn  appearance,  so  that  one  or  two  of  its 
stems  are  abraded  longitudinally  on  the  outer  side,  it  looks  as  if  the  whole  "  trunk  " 
approximated  closely  to  the  original  dimensions  for  the  reasons  given  on  p.  172. 
The  increased  abundance  of  roots  near  the  base  and  the  failure  of  preservation  of 
other  detail  in  the  basal  part  suggests  that  this  is  near  to  the  true  base  of  the  speci- 
men which,  must,  therefore,  have  been  on  a  much  smaller  scale  than  other  known 
species. 

Summarizing  points  of  resemblance,  some  of  which  have  been  mentioned  in  the 
description,  the  Kent  Tempskya  and  T.  grandis  have  in  common:  xylem  containing 
an  appreciable  amount  of  parenchyma;  an  inner  parenchymatous  cortex  which 
encloses  a  constant  but  narrow  irregular  band  of  sclerenchyma  close  to  but  not 


178  A   NEW   TEMPSKYA    FROM   KENT 

contiguous  with  the  stele;  a  similar  narrow  zone  of  parenchyma  forming  the  outer 
layers  of  pith  in  contiguity  with  the  stele  and  with  the  sclerenchyma  inside;  large 
individual  stems  in  the  false  "  stem  "  with  rather  short  internodes.  Thus  apart 
from  difference  of  size  which  may  not  necessarily  be  of  great  significance  the  material 
of  the  Kent  specimen  can  only  be  distinguished  from  Tempskya  grandis  by  the 
smaller  number  of  meristeles  in  the  transverse  sections  of  the  rhizomes.  It  has  from 
one  to  three,  commonly  two  meristeles,  whereas  in  T.  grandis  there  are  two  to  five, 
commonly  three  or  four,  implying  greater  crowding  of  the  leaves  in  this  latter  species 
(cf.  Read  &  Brown  1937  :  115,  text-fig.  3,  pi.  28,  fig.  2  [Note  the  erratum  slip]; 
pi.  32,  figs.  2-5;   pi.  33,  figs.  1-4). 

T.  superba  described  by  Arnold  (1958  :  138)  has  suffered  obliteration  of  much  de- 
tail by  complete  silicification  of  the  tissues.  Form  and  size  of  individual  stems  are 
retained  and  there  is  enough  structure  to  show  a  strong  resemblance  to  T.  grandis 
in  that  both  have  the  distinctive  continuous  but  irregular  sclerotic  layer  in  the  inner 
cortex  absent  in  other  species  except  the  Kent  form.  All  three  also  have  scleren-. 
chyma  in  the  pith.  Arnold  separates  T.  superba  from  T.  grandis  partly  on  the  much 
larger  size  of  its  stems  which  he  gives  as  at  least  1  cm.  in  diameter  without  attached 
leaf  bases,  2  cm.  if  they  be  included.  He  infers  that  the  diameter  of  T.  grandis 
(stems)  is  6-7  mm.  but  according  to  measurements  based  on  Read  &  Brown's 
figures  the  two  species  appear  to  approach  one  another  in  this  respect,  while  in  the 
Kent  Tempskya  a  considerable  range  of  size  occurs  in  the  one  specimen  depending 
on  the  position  within  the  length  of  the  stem  (cf.  p.  172).  Size  of  stem  alone,  there- 
fore, does  not  appear  to  be  a  sufficient  reason  for  separation.  The  holotype  of 
T.  grandis  is  8  cm.  in  diameter,  20  cm.  long.  That  of  T.  superba  was  a  slab  measur- 
ing about  6  by  12  cm.,  2  cm.  thick  before  cutting.  It  was  obviously  very  incomplete. 
Both  greatly  exceed  the  dimensions  given  for  the  Kent  specimen  (see  p.  172). 
T.  superba  normally  shows  four  or  five  foliar  traces  (meristeles)  in  each  stem  section, 
indicating,  as  in  T.  grandis  exceptionally  short  internodes.  On  the  grounds  of 
number  of  foliar  traces  it  therefore  appears  that  the  Tempskya  from  Kent  is  distinct 
from  these  two  species  which  otherwise  it  closely  resembles  in  detailed  cytology. 
T.  superba  was  found  in  an  eroded  Oligocene  deposit  in  Nebraska  but  is  believed  to 
have  been  derived  from  the  Lower  Cretaceous  Dakota  Sandstone. 

As  only  one  specimen  from  Kent  is  in  existence  it  is  merely  described  as  Temp- 
skya sp.  It  is  the  second  species  to  have  been  discovered  in  England.  A  much 
earlier  record  is  T.  erosa  Stokes,  Webb  &  Mantell  from  the  Wealden  Beds  or  Lower 
Greensand  of  Tilgate  Forest,  Hastings  and  Potton.  T.  erosa  differs  completely  in  its 
character  and  preservation  from  Tempskya  sp.  described  here.  Its  numerous  much 
smaller  stems  show  little  structure  in  detail  but  are  embedded  in  dense  masses  of 
innumerable  roots.  It  has  been  re-described  and  discussed  by  Stopes  (1915  :  16) 
and  more  recently  illustrated  by  Seward  (1924,  pi.  16,  fig.  4;  pi.  17,  fig.  16).  It  is 
so  different  in  every  way  that  it  does  not  concern  us  here  any  further. 

Origin  of  the  Specimen.  As  soon  as  the  specimen  was  referred  to  Tempskya  it 
raised  the  question  whether  a  supposed  Mesozoic  form  found  at  Sheppey  could  have 
come  from  the  London  Clay,  especially  in  view  of  the  silicious  preservation  (see 
p.  171).     Mr.  G.  F.  Elliot  has  pointed  out  that  a  similar  preservation  is  known  in  Palm 


A   NEW   TEMPSKYA    FROM   KENT  179 

and  dicotyledonous  wood,  in  Osmunda  dowkeri  and  in  cones  of  Pinns  macrocephalus 
from  Heme  Bay.  Pinus  macrocephalus  is  occasionally  picked  up  on  the  shore 
between  Bishopstone  and  Reculvers  where  the  Thanetian  outcrop  is  exposed.  Some 
of  the  above  have  been  attributed  to  the  Thanet  Sands,  others  to  the  London  Clay  but 
their  preservation  is  different  from  that  of  Tempskya  except  in  the  case  of  the 
dicotyledonous  woods  from  Heme  Bay.  A  few  of  these  (B.M.N.H.,  ¥.27923)  which 
have  closely  comparable  and  characteristic  preservation  have  come  from  unweathered 
foreshore  outcrops  of  the  Woolwich  Bottom  Bed  in  Heme  Bay.  Material  washed  out 
of  the  Thanet  Sands  has  been  found  well  to  the  west  of  Heme  Bay  hence  Tempskya 
could  have  been  transported  naturally  to  the  beach  at  Sheppey.  To  summarize,  the 
preservation  suggests  that  the  Lower  part  of  the  Woolwich  Beds  at  Heme  Bay  is  the 
most  likely  source  of  the  specimen.  The  difficulty  of  accepting  such  a  source  is 
eliminated  if,  as  seems  probable,  such  a  habit  of  growth  is  not  confined  to  the  Mesozoic 
The  photographs  were  taken  in  the  Photographic  Department  of  the  British 
Museum  (Natural  History).  The  typing  was  done  by  Mrs.  M.  Firth.  To  all  the 
persons  concerned  and  to  those  already  mentioned  in  the  text  the  author's  warmest 
thanks  are  given. 

REFERENCES 

Andrews,  H.  N.     1961.     Studies  in  Paleobotany .    xii  +  487  pp.  New  York. 

Andrews,  H.  N.  &  Kern,  E.  M.      1947.     The  Idaho  Tempskya  and  Associated  Fossil  Plants. 

Ann.  Mo.  bot.  Gdn.,  34  :  1 19-186,  pis.  15-27. 
Arnold,   C.   A.     1958.     A  new  Tempskya.   Contr.  Mus.  Paleont.    Univ.  Mich.,   14  :  133-142, 

3  pis. 
Corda,  A.  J.     1845.     Flora  protogaea:  Beitrage  zur  Flora  der  Vorwelt.    128  pp.,  50  pis.  Prague. 
Feistmantel,    O.     1872.     Uber   Baumfarrenreste   der   bohmischen    Steinkohlen,    Perm  -  und 

Kreide-formation.     K.  bohm.  Gesell.  Wiss.  Abh.,  6  :  1-30,  pis.  1,  2. 
Kidston,  R.  &  Gwynne-Vaughan,  D.  T.    1 91 1.     On  a  new  species  of  Tempskya  from  Russia. 

Russ.  K.  min.  Gesell.   Verh.,  48  :  1-20,  pis.  1-3. 
Read,  C.  B.  &  Brown,  R.  W.     1937.     American  Cretaceous  Ferns  of  the  genus  Tempskya. 

Prof.  Pap.  U.S.  geol.  Surv.,  186  :  105-131,  pis.  28-43. 
Seward,  A.  C.     1924.     On  a  new  species  of  Tempskya  from  Montana:    Tempskya  Knowltoni 

sp.  nov.  Ann.  Bot.,  38  :  485-507,  pis.  16,  17. 
Stopes,  M.  C.     1915.     Catalogue  of  the  Mesozoic  Plants  in  the  British  Museum  {Natural  History). 

The  Cretaceous  Flora.     Part  II.     Lower  Greensand  (Aptian)  Plants  of  Britain,     xxxvi  -+- 

360  pp.  32  pis. 

DESCRIPTION   OF   PLATES 

Note.  The  photographer  has  taken  PL  1,  fig.  3  and  Pis.  3-5  from  the  back  of  the 
slide.  Allowance  must  be  made  for  this  mirror  image  when  examining  the  plates  in 
order  to  study  the  development  of  the  stele  and  leaf  traces  in  the  rhizomes. 


PLATE    i 
Tempskya  sp. 

Figs,  i,  2.  "  False  stem  "  from  opposite  sides.  The  apex  has  been  removed.  V.51841, 
upper  surface  of  fragment  so  numbered.  Position  of  slides  is  at  /,  g-i.  On  the  left  in  Fig.  1 
dichotomous  forking  of  the  small  stems  is  seen.  The  white  hollow  at  base  shows  scales  in  the 
actual  specimen  and  upwardly  directed  roots.  In  Fig.  2  rounded  projections  at  p.  are  much 
abraded  petiole  bases.      X  1  approx. 

Fig.  3.  Transverse  section  of  "  stem"  somewhat  crushed  and  disorganized  in  the  lower  half. 
Seven  numbered  true  stems  are  visible  (part  only  preserved  of  1  and  3),  2  and  7  are  embedded  in 
roots  and  scales  of  "  stem  "  so  do  not  in  this  section  touch  the  surface  of  the  specimen.  4  is 
in  process  of  dichotomous  division,  5  and  6  are  rather  poorly  preserved.      X  2.     V. 51 841/. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 

I 


PLATE  1 


■!'    k  . v-5 
•i 


841 


.-* 


GEOL.  15,  4 


19 


PLATE    2 
Tempskya  sp. 

Fig.  4.  Side  of  severed  apical  fragment  showing  projecting  petioles  at  p.  Fragment  cut 
longitudinally  on  right  in  preparing  other  sections  but  here  shows  only  roots  and  scales  out  of 
focus.      X2.     ¥.518416. 

Fig.  5.  Transverse  section  of  typical  diarch  root  penetrating  scales  arising  from  the  epidermis 
of  a  stem.  Large  metaxylem  tracheids  occupy  centre  of  root  and  are  flanked  by  two  patches,  at 
opposite  poles,  of  small  protoxylem  tracheids.  Tissues  outside  xylem,  including  endodermal 
ring,  decayed  (white  in  figure).  Next  come  three  concentric  and  radial  rows  of  sclerenchyma, 
sc,  with  cavities.  Outside  again  are  four  to  five  rows  of  dense  sclerenchyma  cells,  d,  with 
blocked  cavities.  A  diamond-shaped  area  (left  half  only  preserved)  of  thin-walled  cells  beyond 
the  sclerenchyma  is  cortex,  0.  The  multicellular  character  of  scales,  s,  can  be  seen  obscurely 
(out  of  focus).      Xcago.     ¥.518412. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  2 


SK-- ',;.«,  ^  .  v..  - 


•^-a 


**B 


i/J"         '•".$   ;/-x 


fete- 


Sfe...      .     ■fir''"'*"-  '  '^  ^•^Sr^Pfii. '  *"*  "2   '"'■'•  ^*  v*r> 


PLATE   3 

Tempskya  sp. 

Fig.  6.  Part  of  slide  V .  5 1841/ shown  in  Plate  1,  fig.  3.  The  mass  of  roots  and  scales  in  which 
the  stems  are  embedded  are  sectioned  in  various  directions.  Stem  b  (2  in  PI.  I,  fig.  3)  shows  a 
typical  cylindrical  stele  with  protuberance  on  the  left  marking  initiation  of  a  leaf  trace.  De- 
tached meristeles  of  two  incipient  leaf-bases  are  also  seen.  Stem  a,  (7  in  PI.  1,  fig.  3)  shows  a 
stele  with  two  leaf  gaps  one  each  side  of  a  small  island  of  xylem.  Two  curved  arms  of  stele 
indicate  leaf  traces  not  yet  severed.  A  stem  at  c  (4  in  PI.  1,  fig.  3)  which  has  begun  to  divide 
dichotomously  is,  in  consequence,  bilobed.  It  is  outlined  in  white  but  is  shown  untouched  and 
more  highly  magnified  in  PI.  4,  fig.  7.  The  obscure  appearance  is  caused  by  the  partial  decay  and 
dislocation  on  this  side  of  the  "  stem  "  but  the  two  new  steles  are  already  separated.      X  5-5. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  3 


PLATE   4 

Tempskya  sp. 

Fig.  7.  The  dividing  stem  in  PI.  3,  fig.  6  (the  right  margin  of  the  figure  just  cuts  the  edge  of 
the  stem).  Originally  a  cylinder  the  stele  on  the  right  has  been  crushed  and  consequently  dis- 
located in  four  places.  One  break  passes  through  the  prominence  on  the  left  which  indicates  the 
beginnings  of  a  leaf  trace.  A  meristele  which  has  already  separated  is  seen  (surrounded  by 
white)  below  the  stele.  The  second  stele  in  the  left  lobe  of  the  dichotomy  is  also  distorted  but, 
again,  the  lobe  (dislocated)  of  an  incipient  leaf  trace  is  visible  (above  white  patch).  X  10. 
V.51841/. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  4 


PLATE   5 
Tempskya  sp. 

Fig.  8.     The  join  on  the  right  was  due  to  the  stem  lying  in  part  on  two  negatives. 

Stem  a  in  Plate  3,  fig.  6.  Dark  band  at  base  crossing  cortex  and  epidermis  is  a  root.  The  two 
leaf  gaps  flanking  the  island  of  stele  are  better  seen  here  as  are  the  curved  incipient  meristeles 
on  each  side.  The  curved  loop  on  the  left  is  almost  separated  but  that  on  the  right  is  still  fully 
attached.  A  deeply  curved  leaf-trace  (top  left)  is  about  to  emerge.  It  is  already  partially  cut 
off  from  the  parent  stem  by  ingrowing  sclerenchyma  from  the  two  sides.  Both  parent  stele  and 
meristele  are  partly  embraced  externally  by  an  irregular  band  of  sclerenchyma  much  infested 
with  "  foreign  "  roots.  This  belt  of  tissue  is  surrounded  by  parenchyma  on  both  sides.  X  10. 
V.51841/. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  5 


GEOL.  15,  4 


PLATE   6 

Tempskya  sp. 

Fig.  9.  Stems  a,  b  and  in  part  c  show  a  later  (younger)  stage  of  these  stems  in  Plate  3  (re- 
versed as  in  a  mirror).  The  stele  bounding  one  leaf  gap  (in  a)  has  now  united  again  with  the 
central  island  of  xylem  and  forms  the  loop  for  a  new  meristele.  The  meristele  on  the  left  in 
PI.  3  has  separated  and  moved  out  into  the  cortex  (right  in  PI.  6).  The  other  leaf  gap  still 
persists.  In  b  the  bulge  seen  on  the  left  in  PI.  3  has  separated  from  the  stelar  ring  (on  right  in 
PI.  6) ;  the  angular  thickenings  flanking  the  bulge  in  PI.  3  have  elongated  and  united,  but  two 
new  loops  have  formed,  one  on  each  side  of  the  bilobed  fused  xylem  from  which  they  have 
severed  themselves  at  their  upper  ends  thereby  producing  two  curved  loops  and  two  leaf  gaps. 
The  lower  limb  in  PI.  6  has  just  separated  from  the  stem  stele.  In  c  the  formation  of  another 
stele  is  visible.  X5.  V,  51841^. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  6 


PLATE    7 

Tempskya  sp. 

Fig.  io.  A  yet  higher  level  in  stem  a.  The  loop  in  PL  6,  top  left,  has  now  formed  a  new 
meristele.  The  gap  in  the  stele  of  the  stem  still  persists  but  another  loop  at  the  end  of  the  lower 
free  limb  is  about  to  initiate  another  meristele.  The  leaf  trace  (top  right)  is  now  free  from  the 
parent  stem  and  is  surrounded  by  its  own  epidermis.  The  successive  cortical  coats  (see  p.  173) 
are  well  marked.  The  irregular  sclerenchyma  belt  within  the  parenchyma  around  the  stele  is 
little  infested  with  "  foreign  "  roots  here,  less  so  than  in  the  young  meristele  (centre,  above). 
The  scales  arising  from  the  epidermis  are  well  preserved  but  the  magnification  is  insufficient 
to  show  the  multicellular  structure.     X  10.  V.  51841ft. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,4 


PLATE  7 


R*.  M&:JF. 


2E*  t:'.i*> 


PLATE   8 

Tempskya  sp. 

Fig.  ii.  Stem  &  at  a  higher  level  than  in  PI.  6,  and  differently  oriented.  The  meristele  of 
PI.  6  has  separated  from  the  stem  and  is  not  shown  in  this  section.  The  bilobed  island  of  stele 
in  PL  6  has  reunited  with  the  arms  of  the  main  stele  from  which  the  incipient  meristeles,  there 
seen  as  curved  extremities,  have  here  separated  and  moved  outwards.  The  upper  of  these  two 
meristeles  is  now  partly  separated  from  the  stem  stele  by  the  ingrowing  of  the  outermost  band 
of  sclerenchyma.  "  Foreign  "  roots  are  clearly  shown  in  the  pith  sclerenchyma  and  in  that 
outside  the  main  stele  itself.  The  narrow  loop  of  a  new  meristele  is  seen  on  the  radius  between 
the  two  separated  meristeles.      X  io.  V.  518411. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  8 


PLATE   9 

Tempskya  sp. 

Fig.  12.  A  higher  section  through  stem  a.  The  loop  in  the  stele  in  PI.  7  has  separated  to  form 
a  meristele  leaving  a  new  leaf  gap  while  the  gap  seen  in  PI.  7  has  closed.  The  curved  hook-like 
free  end  of  the  stem  stele  in  PI.  7  has  passed  out  as  a  meristele  (right,  below)  partially  separated 
by  ingrowing  sclerenchyma.  The  meristele  above  in  PL  7  is  now  almost  separated  from  the 
parent  stem  as  is  evident  from  the  constriction  which  has  formed  on  each  side  of  it  and  the 
thick  sclerenchyma  between  the  two.  The  separated  trace  in  PI.  7  has  grown  right  out  of  the 
"  false  stem  ".  The  successive  coats :  epidermis  with  thin  outer  cortex,  sclerenchymatous  middle 
cortex,  thick  parenchymatous  inner  cortex  with  included  irregular  band  of  sclerenchyma  just 
outside  the  stele  but  separated  from  it  by  parenchyma,  are  clearly  seen.      X  10.  ¥.51841?'. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  9 


sgg&w 


12 


PLATE   10 

Tempskya  sp. 

Fig.  13.  An  arc  of  the  stele  in  PI.  8.  It  shows  the  large  metaxylem  tracheids  with  scattered 
patches  of  parenchyma  among  them.  Small  tracheids  are  seen  at  the  lower  angle  of  the  thicken- 
ed part  of  the  stele  above  which,  to  the  right,  transverse  orientation  of  tracheids  is  apparent 
where  a  leaf  gap  is  in  process  of  development.  Parenchyma  cells  with  black  crystalline  in- 
clusions are  seen  in  the  inner  cortex,  and  these  cells  abut  on  the  stele  on  both  sides.  Thick- 
walled  sclerenchyma  with  small  cavities  shows  clearly  in  the  centre  of  the  pith  but  is  largely 
obscured  by  "  foreign  "  roots  in  the  irregular  sclerenchyma  belt  of  the  inner  cortex  outside  the 
stele  and,  in  places,  in  the  pith.      X  35.     V.5i84ii. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  10 


P&SBii 


PLATE   ii 

Tempskya  sp. 

Fig.  14.  Stelar  arc  on  the  opposite  side  of  the  stele  in  PI.  8.  Sclerenchyma  cells  of  the 
irregular  band  are  seen  at  sc  in  the  parenchymatous  inner  cortex  where  "  foreign  "  roots  are 
absent.  Small  marginal  tracheids  are  well  developed  in  the  south-east  corner  of  the  stele  and 
especially  where  roots  are  in  process  of  formation.  Tangential  cells  associated  with  endodermis 
are  visible  on  the  right.     Other  features  as  in  PI.  10.      X35.     V. 518412'. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  11 


14 


PLATE    12 

Tempskya  sp. 

Fig.  15.  An  arc  of  stele  in  PL  9,  stem  a,  showing  the  origin  of  two  roots.  Scales  arise  from  the 
epidermis  (top  left)  where  also  a  large  root  emerges  from  the  stem.  Dense  outer  cortex  is 
visible  (top  left)  sharply  differentiated  from  the  sclerenchyma  of  the  middle  cortex  which  in  its 
turn  is  distinct  from  the  parenchyma  with  black  crystalline  inclusions  of  the  inner  cortex. 
The  irregular  sclerenchyma  belt  of  the  inner  cortex  lies  outside  the  stele  but  separated  from 
it  by  a  thin  layer  of  the  parenchyma.  The  endodermis  is  somewhat  blurred  owing  to  the  slight 
obliquity  of  the  section  but  can  be  seen  in  places  in  the  slide  by  focusing.  Associated  tangen- 
tially  elongate  cells  are  clear  in  the  slide,  obscurely  seen  in  the  figure  (right,  inside  stele).  X  35. 
V. 51841*. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  4 


PLATE  12 


15 


12  JAN  1968 


PRINTED  IN  GREAT  BRITAIN 
BY  ADLARD  &  SON  LIMITED 
BARTHOLOMEW   PRESS,   DORKING 


COLONIAL   PHILLIPSASTRAEIDA* 

FROM  THE  DEVONIAN  OF 
SOUTH-EAST  DEVON,  ENGLAND 


C.  T.  SCRUTTON 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  5 

LONDON:  1968 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  THE# 
DEVONIAN  OF  SOUTH-EAST  DEVON,  ENGLAND5*' 


%       J 


BY 

COLIN  THOMAS  SCRUTTON 


-*' 


Pp.  181-281;    18  Plates;   21  Text-figures 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  5 

LONDON:  1968 


THE  BULLETIN  OF  THE  BRITISH  MUSEUM 
(NATURAL  HISTORY),  instituted  in  1949,  is 
issued  in  five  series  corresponding  to  the  Departments 
of  the  Museum,  and  an  Historical  series. 

Parts  will  appear  at  irregular  intervals  as  they  become 
ready.  Volumes  will  contain  about  three  or  four 
hundred  pages,  and  will  not  necessarily  be  completed 
within    one    calendar  year. 

In  1965  a  separate  supplementary  series  of  longer 
papers  was  instituted,  numbered  serially  for  each 
Department. 

This  paper  is  Vol.  15,  No.  5  of  the  Geological 
(Palaeontological)  series.  The  abbreviated  titles  of 
periodicals  cited  follow  those  of  the  World  List  of 
Scientific  Periodicals. 


World  List   abbreviation: 
Bull.  Br.  Mus.  nat.  Hist.  (Geol.) 


Trustees  of  the  British  Museum  (Natural  History)  1968 


TRUSTEES    OF 
THE    BRITISH    MUSEUM    (NATURAL    HISTORY) 

Issued  19  March,  1968  Price  £4 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  THE 
DEVONIAN  OF  SOUTH-EAST  DEVON,  ENGLAND 

By  C.  T.  SCRUTTON 

MS  accepted  April  26th  1967 


CONTENTS 

Page 

I. 

Introduction          ......... 

184 

11. 

Acknowledgments           ........ 

184 

III. 

Stratigraphy           ......... 

185 

(a)  Middle  Devonian      ........ 

185 

(b)  Upper  Devonian       ........ 

188 

IV. 

General  Palaeontology  of  the  Colonial  Phillipsastraeidae 

190 

(a)  Microstructure           ........ 

190 

(b)  Increase.          ......... 

193 

(c)  Variation         ......... 

198 

(i)  Introduction    ....... 

198 

(ii)  Variation  in  diameters       ..... 

200 

(iii)  Variation  in  septal  number  and  the  septal  ratios 

203 

(iv)  Variation  in  size  ratios      ..... 

205 

V. 

Systematic  Descriptions         ...... 

205 

Family  Phillipsastraeidae  C.  F.  Roemer  .... 

208 

Genus  Phillipsastrea  d'Orbigny    ..... 

210 

P.  hennahi  hennahi  (Lonsdale)    .... 

214 

P.  hennahi  ussheri  subsp.  nov.    .... 

221 

P.  devoniensis  (Edwards  &  Haime) 

226 

P.  ananas  (Goldfuss)          ..... 

228 

P.  rozkowskae  sp.  nov.        ..... 

230 

Genus  Frechastraea  nov.      ...... 

231 

F.  pentagona  pentagona  (Goldfuss) 

233 

F.  pentagona  minima  (Rozkowska) 

236 

F.  micrommata  (C.  F.  Roemer)  .... 

240 

F.  carinata  sp.  nov.  ...... 

242 

F.  goldfussi  (de  Verneuil  &  Haime) 

247 

F.  bowerbanki  (Edwards  &  Haime) 

253 

Genus  Thamnophyllum  Penecke  ..... 

257 

T.  germanicum  germanicum  nom.  nov. 

260 

T.  germanicum  schouppei  nom.  nov.    . 

261 

T.  caespitosum  (Goldfuss)  sensu  lato    . 

265 

T.  caespitosum  paucitabulatum  subsp.  nov. 

267 

Thamnophyllum  spp.          ..... 

271 

Genus  Peneckiella  Soshkina           ..... 

271 

P.  salternensis  sp.  nov.       ..... 

273 

VI. 

References   ......... 

275 

SYNOPSIS 
The  classification  of  the  family  Phillipsastraeidae  Roemer  is  critically  reviewed.     Twelve 
species  belonging  to  the  genera  Phillipsastrea,  Frechastraea  gen.    nov.,  Thamnophyllum  and 
Peneckiella,  and  including  Phillipsastrea  hennahi  ussheri  subsp.  nov.,  P.  rozkowskae  sp.  nov., 

GEOL.  15,  5.  21 


184  COLONIAL  PHILLIPS ASTRAEIDAE  FROM  S.E.  DEVON 

Frechastraea  carinata  sp.  nov.,  Thamnophylhim  caespitosum  paucitabulatum  subsp.  nov.  and 
Peneckiella  salternensis  sp.  nov.  are  described  from  the  Middle  and  Upper  Devonian  of  the 
Torquay,  Paignton  and  Newton  Abbot  areas  of  south-east  Devon.  New  names  are  proposed  for 
Thamnophyllum  trigeminum  Penecke  and  Macgeea  [Thamnophylhim)  minima  Schouppe.  De- 
tailed statistical  studies  of  variation  in  species  and  subspecies  samples  and  individual  colonies 
of  many  of  the  taxa  are  described  and  analysed.  The  stratigraphy  of  the  more  important 
localities  from  which  phillipsastraeids  have  been  collected  is  briefly  reviewed. 

I.    INTRODUCTION 

Not  until  the  work  of  Schouppe  (1958),  over  one  hundred  years  after  Lonsdale 
(1840)  had  erected  "  Astrea  hennahii  ",  was  the  presence  of  horseshoe  dissepiments 
in  this,  the  type  species  of  the  genus  Phillipsastrea  generally  accepted.  An  earlier 
record  of  this  fact  by  Smith  (1945  :  37),  who  placed  very  little  emphasis  upon  it, 
appears  to  have  passed  without  notice.  Schouppe,  however,  considered  dissepimental 
form  and  the  associated  trabecular  structure  to  be  of  particular  taxonomic  importance 
in  the  group  of  Devonian  rugose  corals  with  which  Phillipsastrea  is  associated.  He 
thus  attempted  a  thorough  revision  of  their  classification,  but  his  suggestions  are, 
in  part,  unacceptable  where  they  bring  together  species  with  markedly  different 
morphological  characteristics,  unlikely  to  have  been  closely  related. 

The  type  locality  of  "  Astrea  hennahii  "  is  Barton  Quarry,  Torquay  (south  Devon) 
and  many  of  the  colonial  species  and  genera  closely  related  to  Phillipsastrea  are 
also  well  represented  in  this  area.  These  English  corals  have  not  been  examined  in 
detail  since  the  middle  of  the  last  century,  when  Edwards  &  Haime  (1853)  described 
them,  and  their  taxonomic  revision  is  long  overdue. 

In  the  present  paper  the  classification  of  the  phillipsastreids  is  critically  surveyed 
and  the  colonial  species  and  genera  of  the  Phillipsastraeidae  from  south-east  Devon 
are  described.  As  far  as  the  material  allows,  the  variation  in  the  taxa  described  has 
been  investigated  statistically.  Data  collected  for  individual  colonies  of  most  of 
the  species  and  subspecies  enable,  in  addition,  some  comparisons  to  be  made  between 
specific  and  colonial  variation. 

In  the  course  of  this  work  it  has  been  necessary  to  examine  a  number  of  species 
formerly  considered  as  phillipsastreids,  but  which  are  now  removed  from  the 
Phillipsastraeidae.  These  corals  all  belong  to  the  family  Marisastridae  and  have 
been  described  elsewhere  (Scrutton  1967). 

The  following  abbreviations  are  used :  BM(NH)  British  Museum  (Natural  History)  ; 
OUM  University  Museum,  Oxford;  GSM  Geological  Survey  Museum,  London; 
TM  Torquay  Museum;  TM(JB)  Jukes-Browne  Collection  in  the  Torquay  Museum; 
GVM  Dr.  G.  V.  Middleton's  Collection,  Murchison  Museum,  Imperial  College,  London. 

II.  ACKNOWLEDGMENTS 

This  work  formed  part  of  a  thesis  submitted  for  the  degree  of  D.Phil,  at  the 
University  of  Oxford.  The  author  is  indebted  to  both  the  Trustees  of  the  Durham 
Colleges  Research  Studentships  (Durham  University)  and  the  Burdett-Coutts 
Research  Studentships  (Oxford  University)  for  financial  support. 

The  author  is  particularly  grateful  to  Professor  M.  R.  House  (Hull)  who  supervised 
the  work  and  to  Professor  Maria  Rozkowska  (Poznan,  Poland)  and  to  the  late 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  185 

Dr.  H.  Dighton  Thomas  (British  Museum  (Nat.  Hist.)),  for  invaluable  discussion 
on  all  aspects  of  this  paper.  Acknowledgment  is  due  to  the  curators  of  the  Museums 
mentioned  in  the  text,  all  of  whom  have  been  most  helpful  in  making  available 
specimens  in  their  care.  The  photographs  were  taken  by  Mr.  Peter  Green  at  the 
British  Museum  (Nat.  Hist.).  Mrs.  S.  A.  Malcolm  and  Mr.  R.  F.  Wise  (British 
Museum  (Nat.  Hist.))  have  both  given  valuable  technical  assistance  in  the  preparation 
of  this  paper. 

Mr.  M.  Mitchell  (Institute  of  Geological  Sciences,  London)  and  Dr.  G.  F.  Elliott 
have  kindly  read  and  criticized  the  palaeontological  and  stratigraphical  sections  of 
the  manuscript  respectively. 

III.   STRATIGRAPHY 

A  detailed  account  of  the  Devonian  stratigraphy  of  south-east  Devon  is  outside 
the  scope  of  this  work.  As  the  vast  majority  of  the  material  described  comes  from 
only  six  localities,  however,  it  is  intended  to  give  a  brief  account  of  them  here. 
They  are  confined  to  the  limestones  in  the  Torquay,  Paignton  and  Newton  Abbot 
areas  (see  Text-figs.  1,  2).  Further  general  information  on  the  geology  of  south-east 
Devon  can  be  obtained  from  the  Geological  Survey  Memoirs  (Ussher  1903;  Ussher 
et  al.  1913;   Lloyd  1933). 

(a)  Middle  Devonian. 

Dyer's  Quarry  (SX  92206280)  is  situated  in  the  coastal  cliffs  at  the  western  end 
of  the  Daddy  Hole  limestone  mass.  Exposed  to  sea  level  in  the  floor  of  the  disused 
quarry  is  a  15  ft.  sequence  of  thin-bedded,  black,  crinoidal  limestones  exceedingly 
rich  in  coral  remains,  overlain  by  virtually  unfossiliferous  limestones  becoming 
lighter  in  colour  and  more  massive  towards  the  top  of  the  quarry.  The  total  thick- 
ness of  exposed  rock  is  about  60  to  70  ft. 

The  coral  fauna  in  the  floor  of  the  quarry  was  mentioned  briefly  by  Scrutton 
(1965  :  186)  who  suggested  for  it  a  lower  Middle  Givetian  age.  It  is  dominated  by 
colonies  of  Thamnophyllum  germanicum  schouppei  nom.  nov.  (see  p.  120),  relatively 
unbroken  and  apparently  preserved  in  their  position  of  growth.  Some  of  the  simple 
corals  also  appear  to  retain  their  growth  orientation.  This  suggests  a  sheltered 
environment,  either  protected  from,  or  situated  below  the  effects  of  strong  wave 
action,  as  the  large  slender  branching  colonies  of  Thamnophyllum  must  have  been 
rather  delicate  structures  during  life.  The  horizon  with  abundant  Thamnophyllum 
is  sharply  succeeded  by  dark  limestones  with  few  solitary  corals.  Just  above  the 
junction  is  a  thin  band  containing  rounded  limestone  pebbles  which  probably  repre- 
sents contemporaneous  erosion  of  the  sea  floor  by  wave  action. 

Wolborough  Quarry  (SX  85237042),  in  the  south-west  outskirts  of  Newton 
Abbot,  has  produced  a  large  and  varied  fauna  of  Middle  Devonian  aspect  in  the  past 
(Whidborne  1888-1907;  Ussher  et  al.  1913  :  22-24)  although  corals  have  received 
little  mention.  Today,  however,  it  is  badly  overgrown  and  the  accessible  outcrops 
yield  only  scattered  fossils.  The  quarry  is  cut  in  massive,  irregularly  jointed  lime- 
stones, usually  coarsely  crystalline,  with  a  considerable  bioclastic  content  and 
variously  coloured  from  dark  grey  to  a  very  pale  yellowish  tint. 


1 86 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


Phillipsastrea  devoniensis  (Edwards  &  Haime)  has  been  recorded  from  Wolborough 
and  Billingsastraea?  battersbyi  (Edwards  &  Haime)  (see  Scrutton  1967  :  277)  seems 
to  have  been  fairly  common  there  in  the  past.  Phillipsastrea  hennahi  hennahi,  a 
single  specimen,  and  Thamnophyllum  caespitosum  (Goldfuss)  have  now  been  collected 
from  the  quarry  together  with  scattered  solitary  corals  including  Stringophyllum  sp. 

The  exact  stratigraphical  position  of  the  Wolborough  limestone  was  held  to  be 
uncertain  by  Ussher  et  al.  (1913),  but  House  (1963  :  5)  has  shown  that  the  Maenio- 


Fig.  I.     Geological  map  of  south-east  Devon  showing  the  main  collection  sites  thus  +. 
Inset  map  shows  the  position  of  the  area  (based  on  Geological  Survey  maps). 


ceras  molarium  Zone  of  the  middle  Givetian  is  well  developed  here.  Middleton 
(1959)  unfortunately  did  not  describe  Wolborough  representatives  in  his  paper  on 
south  Devon  tetracorals,  but  later  (1960,  table  1)  he  gave  a  general  stratigraphical 
table  for  the  Newton  Abbot  area  based  on  a  modified  version  of  Wedekind's  coral 
zones.  It  appears  that  the  Wolborough  limestone  belongs  to  Middleton's  Givetian 
"  biostromal  and  clastic  limestones  "  unit,  separated,  at  least  in  part,  from  the 
"  biohermal  Frasnian  limestone  "  of  Ramsleigh  Quarry  and  elsewhere  by  tuffs. 

Lummaton  Hill  (SX  91306645).  Probably  more  has  been  written  about  the 
series  of  quarries  now  within  the  northern  outskirts  of  Torquay,  than  any  other 
Devonian  limestone  exposure  in  the  country.  An  early  account  of  the  lithologies 
and  faunas  in  the  quarries  was  given  by  Jukes-Browne  (1906).     Despite  the  massive 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


187 


nature  of  the  limestones,  and  the  undoubted,  though  obscure,  structural  deformation, 
he  suggested  that  a  definite  lithological  succession  could  be  established.  Most  of  the 
lithological  types — the  massive  stromatoporoid  limestone,  the  grey  shelly  limestone 
of  the  Lummaton  Shell  Bed,  and  the  bioclastic  limestone  of  the  more  northern 
exposures — can  still  be  recognized  today.  The  rich  and  varied  Lummaton  fauna 
derives  mainly  from  the  Shell  Bed  and  was  monographed  by  Whidborne  (1888-1907). 
The  latter  did  not  mention  corals,  but  Jukes-Browne  recorded  Phillipsastrea  hennahi 


STAGES 

GONI  ATITE 
ZONES 

TORQUAY 

NEWTON       ABBOT 

LU 
O- 
Q. 

FAMENNIAN 

OSTRACOD 
SLATE 

CURVISPINA 

FRASNIAN 

HOLZAPFELI 

SALTERN             * 
COVE 
BEDS                 + 

BABBACOMBE          ^» 
SLATES                   g* 

CORDATUM 

LUNULICOSTA 

Z 

LU 

Q 
O 

GIVETIAN 

TEREBRATUM 

TORQUAY 
LIMESTONE 

HOPE'S   ^ 
NOSE           "^ 
BEDS            ~^~~ 

SHALES 
Colteolo 

MEADFOOT 

and 

STADDON 

BEOS 

< 

z 
0 

> 

MOLARIUM 
ROUVILLEI 

Z 
< 

z 
> 

0 

0 

EIFELIAN 

JUGLERI 

LATESEPTATUS 
WENKENBACHI 

LU 
O 

EMSIAN 

ile   horizon 

? 

RED   SLATY 

Pencckrello 

LIMESTONE 

\             BIOHERMAL 
\                                * 
-^>         LST 

Ion    and   Botlon 

TUFFS             > 

limestones 

■^ 

BIOSTROMAL          + 

Quarry   hmeslone 

and 
CLASTIC 

LIMESTONE 

SLATE        ! 
with    fasS,ls    |         LST 

T 

SPILITIC       1 

ANDESITE 

sigh     Qui 


Wolborough     Quor, 


Fig. 


2.     Devonian  successions  for  the  Torquay  area  (modified  after  House  &  Selwood  1965) 
and  the  Newton  Abbot  area  (modified  after  Middleton  i960). 


hennahi  and  Haplothecia  pengellyi  (Edwards  &  Haime)  from  the  bioclastic  limestones 
in  the  northern  end  of  the  western  quarries  and  noted  that  they  did  not  occur  else- 
where at  Lummaton.  Unfortunately,  this  particular  locality  has  been  worked  out 
and  these  corals  can  no  longer  be  collected  here.  They  are  still  to  be  found,  however, 
in  limestones  of  a  similar  lithology  exposed  in  nearby  Barton  Quarry. 

Recent  quarrying  near  the  supposed  strike  of  the  Shell  Bed  has  exposed  a  grey 
limestone  with  scattered  brachiopods,  tabulate  and  simple  rugose  corals,  bryozoa  and 
broken  colonies  of  Thamnophyllum  caespitosum  paucitabulatum  subsp.  nov.  This  is 
the  first  record  of  Thamnophyllum  from  Lummaton.  The  lithology  suggests  that 
the  horizon  is  above  rather  than  below  the  Shell  Bed. 

The  Lummaton  fauna  was  described  by  Kayser  (1889  :  186)  as  indicative  of  the 
upper  beds  of  the  Middle  Devonian.  Recent  work  has  fully  substantiated  this. 
Elliott  (1961  :  256  et  seq.)  commented  on  the  apparent  coexistence  of  Stringocephalus 


188  COLONIAL  PHILLI  PS  ASTR  AEI  D  AE  FROM  S.E.  DEVON 

burtini  and  "  Rhynchonella  "  cuboides,  held  to  be  markers  for  the  Givetian  and 
Frasnian  respectively  on  the  Continent.  This  had  caused  Ussher  et  al.  (1913  :  14) 
to  consider  the  Lummaton  limestones  to  be  transitional  across  the  Middle-Upper 
Devonian  boundary.  Elliott  showed,  however,  that  the  "  cuboides  "  from  the 
Lummaton  Shell  Bed  should  be  compared  with  the  Upper  Givetian  Hypothyridina 
procuboides  of  Torley  (1934)  from  Germany  which  strongly  suggests  a  Middle  Devonian 
age  for  this  horizon.  House  (1963  :  6)  later  confirmed  Elliott's  age  for  the  Lummaton 
Shell  Bed.  He  described  a  goniatite  fauna  from  this  level  which  can  be  correlated 
with  the  terebratum  Zone  of  Upper  Givetian  age. 

Elliott  (1961  :  256)  quoted  the  occurrence  of  Phillipsastrea  hennahi  hennahi  at 
Lummaton  and  Barton  as  support  for  the  presence  of  Frasnian  rocks  above  the 
Givetian  at  both  localities.  Middleton  (1959  :  156),  however,  had  already  recorded 
this  subspecies  from  Middle  Devonian  (?  upper  Sparganophyllum  Zone)  limestones 
near  Dartington  Hall  and  it  is  recorded  here  from  undoubted  middle  Givetian  at 
Wolborough  Quarry.  Furthermore  there  is  good  evidence  to  suggest  an  upper 
Givetian  age  for  the  limestones  at  Barton  from  which  P.  hennahi  hennahi  can  now 
be  collected. 

Barton  Quarry  (SX  91246710),  about  a  quarter  of  a  mile  north  of  Lummaton, 
is  now  used  as  a  caravan  camp.  The  quarry  is  cut  in  massive,  mainly  dark-grey 
coarse  crystalline  bioclastic  limestone  which,  in  the  past,  has  yielded  a  large  and 
varied  fauna  monographed  by  Whidborne  (1888-1907)  and  listed  by  Ussher  et  al. 
(1913  :  26).  The  former  did  not  record  the  corals  but  Ussher  et  al.  (1913  :  25) 
described  finer  textured  parts  of  the  limestone  as  abounding  in  corals  and  stromato- 
poroids.  Most  of  the  corals  now  seen  are  tabulates,  mainly  Thamnopora  and  Alveo- 
lites but  it  is  still  possible  to  collect  a  few  specimens  of  Phillipsastrea  hennahi  hennahi 
and  Haplothecia  pengellyi  together  with  Acanthophyllum  sp.  from  the  western  wall 
of  the  old  quarry. 

Barton  is  the  type  locality  for  P.  hennahi.  The  horizon  has  been  considered 
Frasnian  in  age  based  on  early  Continental  records  of  this  subspecies  from  Upper 
Devonian  rocks.  House  (1963  :  6),  however,  has  identified  Wedekindella  brilonense 
(Kayser)  from  among  Whidborne's  Barton  fauna,  presumably  coming  from  the 
massive  limestones  of  the  quarry.  This  goniatite  suggests  the  terebratum  Zone  and 
thus  an  upper  Givetian  age.  In  addition  Acanthophyllum  has  not  so  far  been 
recorded  above  the  Middle  Devonian.  Thus  the  evidence  supports  an  upper  Givetian 
rather  than  a  Frasnian  age  for  the  Barton  limestones  and  on  the  grounds  of  litholo- 
gical  and  faunal  similarity,  the  beds  that  in  the  past  yielded  P.  hennahi  hennahi  and 
Haplothecia  pengellyi  at  Lummaton  are  probably  of  the  same  age. 

(b)  Upper  Devonian. 

The  disused  Ramsleigh  Quarry  (SX  84417015)  is  situated  about  a  quarter  of  a 
mile  east  of  East  Ogwell,  south-west  of  Newton  Abbot.  Exposed  in  the  quarry 
and  in  the  road  cutting  immediately  to  the  south  are  massive,  dominantly  fine 
grained  limestones,  medium  to  pinky  grey  for  the  most  part  but  with  prominent 
lenses  of  a  salmon  pink  colour. 

The  age  of  the  Ramsleigh  limestones  has  been  considered  as  Frasnian  since  the 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  189 

early  eighteen  eighties  at  least.  Ussher  et  at.  (1913  :  15)  quoted  from  an  unpublished 
manuscript  written  by  Champernowne  in  which  the  latter  considered  that  the 
"  splendid  Ramsleigh  mass  ...  is  precisely  the  counterpart  of  the  marble  masses  of 
the  '  Etage  de  Frasne  '  in  Belgium  ".  A  few  years  after  Champernowne  wrote 
this,  Kayser  (1889  :  186)  correlated  the  Ramsleigh  limestone  with  the  Ibergerkalk 
of  Germany.  In  more  recent  times,  Dineley  &  Rhodes  (1956  :  244)  investigated  a 
conodont  fauna  collected  from  a  pale  limestone  band  somewhat  below  the  level  in 
the  quarry  at  which  most  of  the  massive  corals  are  found.  They  concluded  that  the 
fauna  was  Lower  Frasnian  in  age.  Middleton  (1959)  briefly  described  some  of  the 
corals  found  in  the  quarry.  Under  his  description  of  "  Phillipsastraea  pentagona 
var.  micrommata  "  he  wrote  (p.  157)  that  "  According  to  Rozkowska  this  variety 
is  characteristic  of  the  upper  Frasnian  ".  It  is  presumably  on  this  basis  that  he 
considered  (p.  156)  the  Ramsleigh  limestone  to  be  "  probably  middle  or  upper 
Frasnian  "  in  age.  Rozkowska  (1953),  however,  makes  no  definite  statement  of 
the  stratigraphical  range  of  this  coral,  merely  describing  it  from  the  upper  Frasnian 
of  Poland.  Furthermore,  although  Frechastraea  carinata  sp.  nov.  (=  Phillipsastraea 
pentagona  var.  micrommata  of  authors)  does  occur  at  Ramsleigh,  Middleton  mis- 
identified  his  material  which  should  correctly  be  assigned  to  F.  pentagona  minima. 
House  (1963  :  6),  on  the  evidence  of  ammonoids  collected  by  Shannon  from  Rams- 
leigh Quarry,  inferred  that  the  lunulicosta  Zone,  the  lowest  goniatite  zone  in  the 
Frasnian,  is  represented  by  the  massive  limestones.  Thus  the  weight  of  the  fossil 
evidence  suggests  that  these  beds  are  Lower  Frasnian  in  age. 

Quite  a  large  and  varied  collection  of  massive  corals  has  been  made  from  the 
quarry  and  the  adjacent  road  cutting.  Phillipsastrea  hennahi  ussheri  subsp.  nov., 
P.  ananas  (Goldfuss),  P.  rozkowskae  sp.  nov.,  Frechastraea  pentagona  pentagona 
(Goldfuss),  F.  pentagona  minima,  F.  carinata  sp.  nov.,  F.  goldfussi  (de  Verneuil  & 
Haime)  and  F.  bowerbanki  (Edwards  &  Haime)  are  described  in  this  paper  and 
Haplothecia  ogwellensis  Scrutton  elsewhere  (Scrutton  1967  :  272).  Most  of  the 
material  recently  collected  came  from  the  higher  parts  of  the  old  quarry  face  and 
from  a  series  of  outcrops  in  the  road  cutting,  between  20  and  80  yards  west  of  the 
lane  leading  to  the  quarry. 

Saltern  Cove  (SX  895585),  on  the  shores  of  Tor  Bay,  is  i|  miles  south  of  Paignton. 
Resting  on  the  altered  doleritic  rock,  variously  interpreted  as  a  lava  or  a  sill,  which 
forms  the  southern  horn  of  the  cove,  is  a  sequence  of  shales  and  limestones  of  Upper 
Devonian  age  (see  Lloyd  1933  :  86  et  seq.;  House  1963  :  8;  Scrutton  1965  :  188). 
Immediately  above  the  igneous  rock  is  about  20  ft.  of  thick  bedded  limestone  with 
a  distinctive  band  rich  in  broken  colonies  of  Peneckiella  at  the  base  (SX  89505842). 
This  is  the  "  main  Peneckiella  horizon  "  referred  to  elsewhere  in  this  paper.  Higher 
in  the  succession  intercalations  of  red  shale  become  increasingly  important,  separating 
thinner  beds  of  limestone  which  completely  disappear  some  50  ft.  above  the  base  of 
the  sequence.  The  corals  in  these  thin  bedded  limestones  have  been  briefly  men- 
tioned by  Scrutton  (1965  :  188)  as  indicating  a  Frasnian  age.  With  the  presence 
of  the  holzapfeli  Zone  of  Upper  Frasnian  age  established  in  the  northern  end  of  the 
cove,  the  limestone  horizons  are  probably  within  the  Middle  Frasnian  cordatum  Zone 
as  inferred  by  House  (1963  :  7-8). 


igo  COLONIAL  PHILLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

IV.  GENERAL  PALAEONTOLOGY  OF  THE  COLONIAL 
PHILLIPS  ASTRAEIDAE 

Features  of  the  microstructure,  increase  and  variation  in  the  Phillipsastraeidae 
are  sufficiently  uniform  to  warrant  general  treatment.  By  so  doing  much  repetition 
is  saved  in  the  systematic  descriptions  of  individual  species  and  subspecies.  The 
terminology  used  in  this  and  subsequent  sections  is  that  given  by  Moore,  Hill  & 
Wells  (1956)  unless  otherwise  indicated. 

(a)  Microstructure. 

Slight  recrystallization  or  deformation  of  coral  material  may  easily  obscure  the 
details  of  fine  structure  in  the  skeletal  tissue.  Although  the  preservation  of  the 
English  Devonian  material  is  not  particularly  good,  all  the  species  and  subspecies 
described  here  do  show  some  details  of  their  original  microstructure. 

The  septa  are  characterized  by  a  dark,  irregular  median  line  on  either  side  of 
which  are  fibres  of  crystalline  calcite  (see  for  example  PL  13,  fig.  r).  Where  the 
preservation  is  best,  the  fibres  can  be  seen  arranged  in  paired  tufts  or  fascicles  on 
either  side  of  the  septal  axis.  Each  pair  is  presumably  the  cross-section  of  a  single 
monacanthine  trabecula,  with  the  successive  centres  of  calcification  in  the  fibre 
fascicles,  and  thus  in  the  trabecular  axes,  forming  the  dark  median  line.  It  has  not 
been  possible,  however,  to  distinguish  clearly  the  boundaries  of  individual  trabeculae 
in  cross-section. 

The  structure  is  most  clearly  developed  in  the  dilated  part  of  each  septum,  in  the 
zone  immediately  outside  the  tabularium.  The  dilatation  appears  to  be  the  result 
of  simple  swelling  of  the  trabeculae.  There  is  never  more  than  a  slight  offsetting 
of  the  centres  of  calcification — in  other  words,  very  little  break-up  or  zigzagging  of 
the  dark  median  line — to  suggest  that  the  septa  become  multitrabecular. 

The  arrangement  of  the  trabeculae  in  the  vertical  plane  is  reflected  by  the  direction 
of  divergence  of  the  fibre  fascicles  from  the  median  plane  of  the  septum  in  cross- 
section  (see  Kato  1963,  text-fig.  3).  In  the  phillipsastraeids  described  here,  the 
fibres  can  sometimes  be  clearly  seen  changing  their  attitude  to  face  outwards  at 
either  end  of  the  dilated  part  of  the  septum,  corresponding  to  the  fan-shaped  diverg- 
ence of  the  trabeculae  in  the  septal  plane.  The  point  of  divergence  is  located  more 
or  less  centrally  in  the  zone  of  septal  dilatation. 

The  carination  developed  in  Frechastraea  carinata  sp.  nov.  and  Peneckietta  saltern- 
ensis  sp.  nov.  is  due  to  the  development  of  regularly  spaced,  enlarged  trabeculae. 
Whilst  the  trabeculae  retain  their  linear  arrangement  along  the  septal  axis  the  carinae 
are  yard-arm,  but  they  may  become  offset  on  alternate  sides  of  the  septum,  resulting 
in  xyloid  carination.  The  swollen  trabeculae  are  sometimes  separated  by  clear 
structureless  calcite  as  though  the  trabeculae  had  separated  and  the  septa  become 
discontinuous.  To  what  extent  this  effect  is  the  result  of  recrystallization  is  difficult 
to  ascertain. 

In  longitudinal-section,  the  arrangement  of  the  trabeculae  in  the  septal  plane 
can  often  be  clearly  seen  (Text-figs.  3,  yb).  As  mentioned  above,  the  trabeculae  are 
arranged  in  a  fan  and  this  is  centred  on  the  crest  of  the  dissepimentarium,  usually 
formed  by  horseshoe  or  peneckielloid  (Rozkowska  i960  :  32)  dissepiments.     In  the 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


191 


zone  of  septal  dilatation,  which  is  most  frequently  sectioned  in  a  longitudinal  slice, 
the  trabecular  fan  may  often  appear  symmetrical  about  the  axis  of  divergence. 
When  more  rarely  a  longitudinal-section  is  obtained  of  the  thinner  parts  of  the 
septum,  the  symmetrical  arrangement  is  not  maintained.  Particularly  in  the  species 
of  Phillipsastrea  and  Frechastraea,  the  trabeculae  in  the  dissepimentarium  gradually 
return  to  a  vertical  position  as  the  septa  are  traced  towards  the  periphery  of  the 
corallites.  In  the  tabularium,  septa  are  composed  of  trabeculae  from  the  edge  of 
the  fan,  entering  from  the  dissepimentarium  at  a  very  low  angle,  often  almost 
horizontally.  The  attitude  of  the  trabeculae  appears  always  to  be  normal  to  the 
dissepimental  surface  (Text-fig.  3). 

In  the  case  of  Thamnophyllum  (Text-fig.  76),  the  evidence  suggests  that  the 
trabecular  fan  is  more  nearly  symmetrical  in  the  septum  although,  in  the  major 
septa  particularly,  the  centre  of  divergence  is  closer  to  the  epithecal  than  the  axial 
end.  The  fan  in  species  of  Peneckiella  is  somewhat  less  symmetrical  and  more 
variable  in  shape — a  reflection  of  the  diversification  in  dissepimental  form. 


Fig.  3.  Comparative  longitudinal-sections:  a.  Phillipsastrea  hennahi  hennahi  (OUM 
D520/P2) ;  b.  Frechastraea  pentagona  pentagona  (OUM  D537/P2) ;  c.  Frechastraea  carinata 
(OUM  D3 1  ob).     All      x8. 


192 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


In  the  massive  species  described  here,  the  septa  are  either  more  or  less  confluent 
between  corallites  (e.g.  Frechastraea  bowerbanki)  or  intermingle  to  form  a  pseudo- 
theca.  The  role  of  individual  septa  is  usually  clear  in  the  formation  of  a  diffuse 
and  irregular  pseudotheca  such  as  that  in  the  astraeoid  Phillipsastrea  hennahi  (Text- 
fig.  4c).  In  some  species  of  Frechastraea,  however,  the  peripheral  ends  of  the  septa 
are  sharply  geniculate  and  form  a  very  strong  wall  by  their  fusion  with  one  another 
(Text-fig.  46).  The  part  played  by  the  individual  septa  is  not  clear  although  the 
septal  characteristics  of  the  pseudotheca  are  obvious.  In  Frechastraea  carinata, 
carinae  may  be  rarely  seen  on  the  pseudotheca  where  the  septal  carination  has  been 
carried  over  into  the  wall.  In  the  past,  massive  corals  with  such  walls  have  been 
called  cerioid  (Rozkowska  1953  :  62)  but  this  term  should  be  strictly  confined  to 
massive  corals  in  which  an  epitheca  still  surrounds  individual  corallites  (Text-fig.  \a) 
(Lang  1923  :  123;  Hill  1935  :  488).  Thus  the  term  "  pseudocerioid  "  is  introduced 
here  to  describe  such  corals  as  Frechastraea  pentagona  pentagona,  F.  carinata  and  F. 
goldfussi  in  which  a  strong  pseudotheca  is  built  up  by  modified  septal  elements.1 

1  See  Addendum. 


Fig.  4.  Comparative  wall  structures:  a.  cerioid — Hexagonaria  firthi  (BM(NH)  R29476) ; 
b.  pseudocerioid — Frechastraea  goldfussi  (OUM  D539/P2) ;  c.  weak  astraeoid — Phillip- 
sastrea hennahi  hennahi  (OUM  D520/P  1).     All   X5. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.   DEVON 


193 


All  the  massive  coralla  described  here  are  presumably  surrounded  by  a  holotheca 
(Hill  1935  :  497)  although  it  has  only  been  observed  in  Frechastraea  micrommata 
(C.  F.  Roemer)  and  in  two  sections  of  F.  goldfussi.  It  is  fibronormal  in  character 
and  somewhat  variable  in  thickness  in  the  range  o*i-o*i5  mm.  Both  the  septa  and 
the  pseudotheca  abut  with  a  shallow  convex  surface  against  the  holotheca,  or  penetrate 
slightly  to  form  a  wedge-shaped  depression  in  its  surface. 

Thamnophyllum  germanicum  schouppei,  T.  caespitosum,  T.  caespitosum  paucita- 
bulatum  and  Peneckiella  salternensis  have  cylindrical  corallites  with  fibronormal 
epithecae.  The  peripheral  septal  ends  meet  the  epitheca  and  depress  it  slightly 
(Kato  1963,  text-fig.  17/)  in  the  same  manner  as  the  relationship  between  septa  and 
holotheca  in  F.  goldfussi. 

In  all  cases  where  the  microstructure  can  be  distinguished,  the  tissue  of  dissepi- 
ments and  tabulae  is  fibronormal. 

(b)  Increase. 

For  full  details  of  corallum  increase  it  is  necessary  to  cut  serial  sections  but  in  the 
present  material,  this  has  been  possible  only  with  Thamnophyllum  germanicum 
schouppei.  Even  in  this  subspecies,  fracturing  has  so  affected  the  point  of  branching 
that  the  details  of  septal  insertion  are  obscured.  Increase  in  the  other  taxa  described 
here  is  known  only  from  random  sections  cut  through  developing  or  immature 
individuals.  Both  axial  and  lateral  increase  are  recorded  and  in  some  cases  both 
may  occur  within  the  same  colony. 

At  the  present  time  no  detailed  work  has  been  done  on  increase  in  plocoid  rugose 
corals.  Most  of  the  methods  of  increase  observed  in  the  present  material,  however, 
have  been  briefly  described  by  Rozkowska  (1953).  The  most  common  process  is 
the  development  of  one  or  more  new  individuals  in  the  border  area  of  two  or  more 
surrounding  adult  corallites  (Text-figs.  5a,  b).     Rozkowska  (1953  :  39,  71)  called 


Fig.  5.  Increase  in  massive  Phillipsastraeidae :  a.  lateral  (intercalicinal) — Frechastraea 
pentagona  minima  (GSM  PF4031);  b.  lateral — Frechastraea  goldfussi  (BM(NH)  R46370) ; 
c.  axial — Frechastraea  goldfussi  (BM(NH)  R46370).     $a  xio;    5ft,  c  x8. 


194  COLONIAL  PHILLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

this  "  intermural  "  increase  in  the  case  of  her  "  cerioid  "  colonies  and  "  intercalicinal  " 
increase  in  the  forms  she  recognized  as  plocoid.  In  fact  Rozkowska's  use  of  cerioid 
is  equivalent  to  the  term  pseudocerioid  as  used  here,  there  being  no  epitheca  but  a 
pseudotheca,  formed  by  modified  septal  ends  separating  individual  corallites :  these 
colonies  are  also,  therefore,  plocoid.  Intermural  increase  was  defined  by  Hill 
(1935  :  491)  in  terms  of  truly  cerioid  coralla  and  it  was  further  restricted  by  Jull 
(1965  :  206)  to  daughter  corallites  which  appear  to  develop  between  corallite  walls 
without  a  particular  parent.  In  view  of  the  literal  meaning  of  "  intermural ", 
Jull's  definition  is  accepted  here  as  it  is  within  the  original  scope  of  the  term  and  is 
most  suitably  described  by  it.  Rozkowska's  "  intermural  "  increase  (Rozkowska 
1953  :  71  did  mention  that  the  term  was  not  altogether  appropriate)  is  better  con- 
sidered as  a  form  of  lateral  increase.  In  some  instances  (Text-fig.  56),  the  daughter 
corallite  appears  to  develop  in  a  very  similar  manner  to  that  described  as  lateral  in 
the  cerioid  Lithostrotion  cf.  portlocki  by  Jull  (1965,  text-fig.  6(1)).  More  usually  in 
these  pseudocerioid  corals,  however,  the  daughter  corallite  has  no  wall  separating  it 
from  most,  if  not  all,  of  the  surrounding  corallites  during  the  early  stages  of  develop- 
ment and  the  parent  corallite  may  be  very  difficult  to  distinguish  (see  Rozkowska 
1953,  text-fig.  35).  This  latter  situation  seems  not  to  differ  fundamentally  from 
intercalicinal  increase  in  the  other  plocoid  corals  with  weak  or  absent  corallite  walls 
(for  example  Text-fig.  5a  and  Rozkowska  1953,  text-fig.  25).  It  is  proposed,  there- 
fore, to  term  both  the  intermural  and  intercalicinal  increase  of  Rozkowska  (1953) 
as  lateral  increase.  Full  understanding  of  these  processes  in  plocoid  coralla,  however, 
must  await  studies  by  serial  sectioning. 

Examples  of  axial  increase  have  been  seen  much  more  rarely  in  the  massive  corals 
(Text-fig.  5c).  Consequently,  knowledge  of  the  process  is  based  on  very  few  sections 
and  cannot  be  described  in  full.  Increase  is  apparently  effected  by  the  elongation 
and  subsequent  bilobation  of  the  tabularium  with  commensurate  insertion  of  new 
septa.  The  parent  tabularium  finally  divides  into  two  new  individuals  and  their 
full  rings  of  septa  are  completed  in  the  area  of  fission.  In  all  of  the  few  examples 
seen  there  are  never  more  than  two  new  individuals  formed  at  the  same  time. 

Axial  increase  has  been  recorded  in  three  massive  species,  all  of  which  have  pre- 
dominant lateral  increase  (Table  1). 

The  process  interpreted  as  "  axial  increase  "  by  Rozkowska  (1953  :  65,  text-fig. 
39)  in  her  "  Phillip sastraea  pentagona  "  has  also  been  observed  in  several  specimens 
among  the  present  material.  From  Rozkowska's  figure  and  a  consideration  of  the 
English  specimens,  however,  it  is  doubtful  whether  this  is  really  increase  but  rather 
a  form  of  rejuvenescence  in  massive  coralla.  There  is  no  indication  that  this 
phenomenon  ever  leads  to  the  formation  of  two  or  more  individuals.  Confirmation 
of  this  interpretation  must  nevertheless  await  evidence  from  serial  sectioning. 

Mode  of  increase  in  the  massive  corals  is  summarized  below  (Table  1).  It  is 
possible  that  axial  increase  may  be  shown  to  occur  in  other  species  than  those 
indicated  when  more  material  has  been  examined. 

Of  the  phaceloid  corals  described  here,  one  has  exclusively  axial  increase  and  the 
others,  exclusively  lateral. 

Thamnophyllum  caespitosum  and  Peneckiella  saltemensis  have  a  style  of  increase 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


195 


Table  i. — Increase  in  massive  Phillipsastraeidae. 

Lateral      Axial 

Phillipsastrea 

P.  hennahi  hennahi  X 

P.  hennahi  ussheri  x 

P.  devoniensis  No  evidence 

P.  ananas  x  1 

P.  rozkowskae  No  evidence 


Frechastreaea 

F.  pentagona  pentagona 

X 

X 

F.  pentagona  minima 

X 

X 

F .  micrommata 

X 

F.  carinata 

X 

X 

F.  goldfussi 

X 

X 

F.  bowerbanki 

X 

1  After  Freeh,  1885,  pi.  2,  fig.  5. 

similar  to  that  described  as  "  thamnophylloid  lateral  "  by  Rozkowska  (i960  :  31). 
The  daughter  corallite  arises  from  the  dissepimental  tissue  of  the  parent  (Text-figs. 
6,  7).  In  the  early  stages,  the  adult  corallite  becomes  egg-shaped  in  cross-section 
with  the  more  pointed  end  containing  dissepimental  tissue  only,  the  septa  having 
withdrawn  from  the  epitheca  in  this  area.  This  projection  expands  in  size  and  septa 
begin  to  appear  on  the  wall  farthest  from  the  parent.  Some  septa  from  the  parent 
itself  may  extend  slightly  into  the  developing  individual  where  the  two  are  joined 
and  these  appear  eventually  to  contribute  to  the  latter 's  full  complement  of  septa. 
As  the  process  continues,  the  daughter  corallite  grows  more  circular  and  forms  a 
bilobed  complex  with  the  parent.     At  the  same  time,  new  septa  are  inserted  on 


Fig.  6.     Lateral  increase  in  Peneckiella  salternensis :  a.  OUM  D553/P1;  b.  OUM  D547/P1; 
c.  OUM  D547/P3  (same  corallite  as  in  6b);   d.  OUM  D553/PI.      All  X5. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


Fig.    7.     Lateral  increase   in  Thamnophyllum  caespitosum  paucitabulatum: 
R46i62rf;    b.  BM(NH)  1*46163(7.     Both   X3. 


a.    BM(NH) 


both  flanks  of  the  daughter  which  now  possesses  most  adult  characteristics.  During 
the  bilobed  stage,  the  two  calices  are  wholly  or  partially  separated  by  an  irregular 
pseudotheca  formed  by  the  geniculation  and  mutual  interference  of  the  septa  in  the 
waist  of  the  complex.  The  daughter  corallite  may  be  connected  to  the  parent  by 
extra-dissepimental  tissue  before  final  separation  when  the  latter's  epitheca  may 
be  complete.  In  Thamnophyllum  caespitosum  two  daughter  corallites  may  be  pro- 
duced at  the  same  level  but  this  has  not  been  observed  in  Peneckiella  salternensis. 

Thamnophyllum  germanicum  schouppei  on  the  other  hand  displays  exclusively 
axial  increase  similar  to  that  found  in  many  other  species  of  Thamnophyllum.  In 
all  the  specimens  examined,  either  three  or  four  daughter  corallites  are  produced  in 
each  case.  One  specimen,  in  which  increase  is  threefold,  has  been  serial  sectioned 
(Text-fig.  8).  The  parent  corallite  is  about  5  mm.  in  diameter  at  the  inception  of 
increase  which  is  marked  by  a  striking  change  in  skeletal  deposition  in  the  tabularium. 
After  the  last  normal  tabulae  are  laid  down  in  the  parent,  the  dissepimentarium 
continues  to  form  as  usual.  In  the  tabularium,  however,  steeply  inclined  plates  are 
secreted  to  form  a  cone  (Text -fig.  8iii)  which  modifies  upwards  into  a  pyramid  with 
as  many  sides  as  daughters  are  produced  (Text-fig.  8iv,  v).  Upon  these  plates  the 
new  individuals  are  built  up.  As  the  cross-section  of  this  axial  structure  changes  in 
threefold  increase  from  circular  to  triangular,  three  septa,  each  opposite  and  extending 
to  join  a  corner  of  the  triangle,  become  increasingly  strongly  developed.  These 
delimit  the  areas  of  the  new  corallites.  Thus  each  new  corallite  inherits  roughly  a 
third  (or  in  fourfold  increase,  a  quarter)  of  the  mature  septa  and  dissepiments 
of  the  adult.  The  first  tabulae  of  each  daughter  are  deposited  between  the  base 
plate  and  the  inherited  dissepiments.  Meanwhile  the  cross-section  of  the  complex 
becomes  increasingly  trilobed  (or  tetralobed)  and  new  septa  are  inserted  along  the 
inner  margins  of  the  developing  corallites  (Text-fig.  8vi,  vii).     The  formation  of  the 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


197 


\       (vii) 


(vi) 


v  A 


^  (V) 


Fig.   8.     Axial  increase  in  Thamnophyllum   germanicum  schouppei:    longitudinal-section 
after  OUM  D272;    serial  cross-sections  after  OUM  D510/P5-10,  pi2-i6.     Spacing  of 
sections    in    mm:     i(p5) — 0-747 — ii(p6) — 1-126 — iii(p7) — 0-712 — iv(p8) — 0-444 — V(P9) — 
0-330— vi(pio)—o-533—vii(pi2)—o-2o8—viii(pi3)— 0-267— ix(pi4)— 0-495— x(pi5)— 
0-574 — xi(pi6).     All  X5. 


geol.  15,  5. 


ig8  COLONIAL  PHILLIPS ASTRAEIDAE  FROM  S.E.  DEVON 

new  horizontal  structural  elements  proceeds  from  the  periphery  towards  the  axial 
area  of  the  complex  and  by  the  time  the  process  is  close  to  completion,  vertical  growth 
has  passed  the  apex  of  the  base  plates  and  the  daughters  are  in  intimate  contact 
(Text-fig.  8viii).  Subsequently,  the  newly  formed  sections  of  the  dissepimentaria 
acquire  adult  characteristics  with  distinct  traces  of  horseshoe  dissepiments  in  cross- 
section  (Text-fig.  8ix,  x)  and  the  new  corallites  complete  their  epithecae.  Evidently 
no  extra-dissepimental  tissue  (caenogenetic  tissue  of  Soshkina  1953)  is  formed  in 
Thamnophyllum  germanicum  schoupftei,  for  as  soon  as  the  normal  dissepimental  tissue 
is  developed,  the  daughters  become  phaceloid  (Text-fig.  8xi). 

(c)  Variation. 

(i)  Introduction.  Sufficient  material  is  available  of  many  of  the  taxa  described 
here  to  allow  their  variation  to  be  studied  in  some  detail.  The  statistics  are  grouped 
with  the  individual  species  and  subspecies  as  part  of  their  characterization  but  their 
great  interest  is  in  the  general  trends  they  show  which  are  commented  upon  here. 

There  are  certain  problems  in  the  statistical  treatment  of  south  Devon  material. 
Only  one  of  the  samples — Thamnophyllum  germanicum  schouppei  from  Dyer's 
Quarry — is  demonstrably  from  a  population  preserved  more  or  less  in  position  of 
growth.  In  all  other  cases  the  faunas  from  which  collections  have  been  made  have 
apparently  suffered  some  post-mortem  movement,  the  extent  of  which  is  difficult  to 
assess.  The  coral  colonies  are  broken  and  disorientated,  in  most  cases  preserved 
in  massive  limestones  which  have  suffered  more  or  less  from  tectonic  stresses  and 
recryst  allization . 

The  collections  thus  consist  of  fragmentary  coralla  which  preclude  ontogenetic 
study.  The  influence  of  ontogeny  on  the  data  obtained  from  these  corals,  however, 
appears  to  be  small.  Longitudinal-sections  of  corallites  in  the  massive  colonies  show 
tabularium  diameters  to  be  virtually  constant  over  most  of  their  vertical  growth. 
In  this  respect,  they  appear  to  behave  in  the  same  way  as  the  phaceloid  colonies,  in 
which  the  influence  of  ontogeny  is  minimal  in  the  extensive  cylindrical  parts  of  the 
corallites. 

All  measurements  have  been  made  in  cross-sections.  In  many  colonial  corals,  a 
large  number  of  corallites  are  unavoidably  cut  at  varying  degrees  of  obliquity, 
resulting  in  elliptical  sections.  As  the  corallites  in  the  phaceloid  colonies  and  the 
tabularia  in  all  the  colonial  corals  considered  here  are  circular  in  sections  normal  to 
their  axes,  diameter  measurements  have  been  made  along  the  minor  axis  of  the 
ellipse.  Care  has  been  taken  to  separate  ellipticity  due  to  oblique  section  from  that 
resulting  from  crushing  or  tectonic  distortion.  Corallites  which  appear  to  have 
been  deformed  in  this  way  have  not  been  measured. 

The  following  dimensions  were  recorded  for  each  corallite: 
d  corallite  diameter.     Measured  only  in  phaceloid  colonies  and  recorded 

to  the  nearest  o-l  mm. 
dt  tabularium   diameter.    All  of   the   corals   described  here  have   a 

clearly  defined  tabularium  junction:  recorded  to  the  nearest  o-i  mm. 
n  number  of  major  septa.     Counts  of  the  number  of  major  septa  were 

recorded  with  the  corresponding  d  and/or  dt  value. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  199 

The  following  were  calculated  in  the  case  of  all  massive  coralla : 

A  average  corallite  area  in  a  colony.     As  it  is  not  possible  to  measure 

corallite  diameter  in  a  massive  colony,  the  average  corallite  area  in 
a  colony  was  measured  to  facilitate  comparison  of  corallite  and 
tabularium  size.  Values  of  A  were  obtained  by  counting  corallite 
numbers  in  a  cross-sectional  area  measured  by  means  of  a  transparent 
graticule  divided  into  squares  of  0-5  cm.  side.  The  number  of 
corallites  was  then  divided  into  the  corresponding  area  and  the  result 
recorded  in  square  centimetres.  Care  was  taken  to  calculate  A  from 
those  corallites  whose  tabularia  were  also  measured.  This  value  is 
subject  to  some  error  due  to  the  inclusion  in  its  calculation  of  some 
obliquely  sectioned  corallites. 

At  average  tabularium  area  in  a  colony.     Calculated  from  dtx  (the 

mean  tabularium  diameter)  of  each  colony  by  the  formula  for  the 
area  of  a  circle  and  recorded  in  square  centimetres. 
From  the  above  basic  data,  the  following  ratios  were  calculated: 

dt/d  tabularium,  corallite  diameter  ratio.    This  ratio  could  only  be 

calculated  for  phaceloid  colonies. 

n/d  or  n/dt  septal  or  septal-tabularium  ratio  respectively.  The  septal  ratio 
as  normally  applied  to  corals  is  the  number  of  major  septa  divided 
by  the  corresponding  corallite  diameter,  which  in  the  present  work 
could  only  be  calculated  for  the  phaceloid  colonies.  In  the  massive 
corals,  the  ratio  of  the  number  of  major  septa  to  the  tabularium 
diameter,  called  the  septal-tabularium  ratio,  was  calculated.  This 
ratio  behaves  in  a  similar  way  to  the  septal  ratio  but  is  not  directly 
comparable  with  it.  It  will,  of  course,  have  higher  values  than 
corresponding  septal  ratios  and  may  vary  differently  with  size, 
depending  on  the  variation  in  the  dt/d  ratio. 

At/A  tabularium  to  corallite  area  ratio.    This  ratio  is  the  approximate 

counterpart  in  massive  corals  of  the  dt/d  ratio  in  phaceloid  corals. 
The  At/A  ratio,  however,  has  only  been  calculated  as  an  average  for 
each  colony  and  not  for  individual  corallites  because  of  the  difficulty 
of  measuring  accurately  the  area  of  an  irregularly  polygonal  corallite. 
For  the  dimensions  and  ratios  mentioned  above,  the  following  standard  statistics 

were  calculated : 

N  sample  size.     The  number  of  corallites  in  the  sample,  with  the  number 

of  colonies  given  in  parentheses.  In  the  case  of  A,  At,  and  the 
ratio  At/A,  N  is  the  number  of  colonies. 

O.R.  overall  range. 

x  mean  value. 

s  standard  deviation. 

C.V.  coefficient  of  variation. 

S.E.m  standard  error  of  the  mean. 

These  elementary  statistics  are  dealt  with  in  many  books  and  their  application  to 

zoology  and  palaeontology  is  discussed  in  Mayr,  Linsley  &  Usinger  (1953)  and 


2oo  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

Simpson,  Roe  &  Lewontin  (i960).  Reference  has  been  made  to  both  these  texts 
during  the  present  work  but  principally,  the  writer  has  followed  Imbrie  (1956). 
The  latter  gives  a  clear  and  concise  account  of  all  the  biometric  techniques  used  here. 
It  was  found  necessary,  however,  in  view  of  the  large  size  of  most  of  the  samples,  to 
calculate  the  statistics  for  all  characters  except  A,  At  and  At/A  by  grouping  the 
data  in  class  intervals  corresponding  to  the  o-i  mm.  intervals  in  which  the  diameters 
were  measured. 

The  statistics  calculated  for  each  species  and  subspecies  and  the  representative 
colonies  are  given  in  tabular  form  with  the  systematic  descriptions  (Tables  4-16). 
The  colonies  selected  for  individual  treatment  were  those  in  which  the  most  corallites 
could  be  measured.  Usually  it  was  possible  to  use  colonies  with  50  or  more  corallites 
but  in  cases  of  species  with  large  calices,  smaller  numbers  had  to  be  used.  The  colony 
with  the  least  data  is  Colony  1  of  Thamnophyllum  germanicum  schouppei  with  23 
measured  corallites. 

The  data  are  also  illustrated  graphically  (Text-figs.  9-21).  In  the  plots  of  dt  against 
d,  n  against  d  or  dt,  n/d  against  d  and  n/dt  against  dt,  diameter  is  recorded  along 
the  abcissa  to  the  nearest  o*i  mm.  As  stated  above,  the  character  on  the  ordinate 
was  averaged  and  plotted  as  a  single  point  in  each  o-i  mm.  class.  Thus  points  in 
the  middle  of  the  ranges  of  values  on  these  graphs  were  based  on  many  more  observa- 
tions than  those  at  either  end. 

On  all  the  graphs,  the  scatter  of  points  approximated  fairly  closely  to  a  straight 
line.  When  similar  graphs  are  plotted  for  complete  ontogenetic  studies,  it  has  been 
shown  that  the  points  usually  fall  on  a  curve  of  quite  complicated  form  (see 
Voynovskiy-Kriger  1954).  In  the  present  case,  however,  the  influence  of  ontogeny 
is  thought  to  be  slight  and  the  data  representative  of  the  mature  growth  stages. 
Voynovskiy-Kriger's  curves  approximate  very  closely  to  straight  lines  in  maturity 
(his  "mature"  plus  "old"  stages),  as  is  shown  by  the  present  results.  Thus  straight 
lines  have  been  calculated  from  the  data  represented  by  the  scatter  of  points  on  each 
graph.  For  ease  of  comparison,  the  lines  only  are  illustrated  in  the  text-figures. 
After  Imbrie  (1956),  the  reduced  major  axis  was  chosen  as  the  most  suitable  line 
for  problems  of  relative  growth. 

For  each  line,  the  formula  is  given  in  the  data  tables  as  follows : 
r  correlation  coefficient 

«„..,.,  .,   ■    •,      „}  line  of  the  form  y  =  ax  +  b 

b     initial  growth  index    J  J 

Where  statistical  discrimination  has  been  used  between  congeneric  species  and 
subspecies,  the  procedure  followed  is  again  that  detailed  by  Imbrie  (1956). 

(ii)  Variation  in  diameters.  The  assessment  of  size  in  these  corals  is  mainly  through 
the  analysis  of  tabularium  diameters  as  corallite  diameters  can  only  be  measured  in 
the  phaceloid  colonies.  Variation  in  size  may  be  the  result  of  genetic,  ontogenetic 
or  ecological  influences.  Because  of  the  parts  played  by  the  latter  two  factors, 
which  are  often  not  easy  to  assess,  care  must  be  taken  in  the  significance  placed  on 
size  differences  and  relative  variation  between  colonies  and  species.  In  the  present 
case,  ecological  control  is  largely  an  unknown  factor  although  the  influence  of  ontogeny 
may  be  regarded  as  minimal. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  201 

The  variation  in  tabularium  diameter  in  each  of  the  taxa  considered  here  is  fairly 
similar  (see  Table  2).  Values  of  C.V.  range  from  773  for  the  sample  of  Frechastraea 
bowerbanki  to  1479  for  the  sample  of  Thamnophyllum  germanicum  schouppei.  Varia- 
tion in  the  former,  however,  is  almost  certainly  underestimated  as  only  four  incom- 
plete colonies  were  available  for  measurement.  Values  of  C.V.  for  the  other  species 
of  Frechastraea  are  all  about  10.  These  figures  are  close  to  those  obtained  by  Oliver 
(i960  :  83,  table  7)  for  solitary  cylindrical  coral  species  and  much  lower  than  the 
variation  he  found  in  conical  forms. 

From  the  present  data  it  is  impossible  to  say  if  massive  corals  are  more  or  less 
variable  than  phaceloid  forms.  Although  Thamnophyllum  germanicum  schouppei 
has  the  highest  C.V.  value,  those  for  T.  caespitosum  paucitabulatum  and  Peneckiella 
salternensis  are  much  the  same  as  the  values  for  many  of  the  massive  corals.  In 
fact,  the  C.V.  figure  for  the  Barton  Quarry  sample  of  Haplothecia  pengellyi,  a  massive 
marisastrid  (see  Scrutton  1967  :  274,  table  2),  is  1575,  which  is  higher  than  that 
for  T.  germanicum  schouppei.  On  the  other  hand,  there  is  a  general  tendency  for 
C.V.  values  to  be  higher  with  increase  in  mean  tabularium  diameter.  In  the  phace- 
loid corals,  variation  in  corallite  diameters  is  roughly  the  same  from  species  to 
species,  although  always  lower,  than  in  the  corresponding  tabularium  diameters. 
Oliver's  results  with  six  solitary  corals  do  not  show  such  consistency  and  tabularium 
diameters  are  less  variable  than  corallite  diameters  in  two  instances. 

Variation  in  size  within  colonies,  with  only  two  exceptions,  was  found  to  be  less 
than  that  in  the  total  samples  of  the  same  species  or  subspecies  from  the  same  locality. 
One  exception  is  Colony  1  of  Frechastraea  bowerbanki  (Table  12)  which  species  has 
already  been  explained  to  be  probably  inadequately  sampled.  The  other  exception 
is  Colony  1  of  Phillipsastrea  hennahi  hennahi  from  Lummaton  (Table  4).  In  this 
case,  the  C.V.  value  for  dt  in  the  colony  only  slightly  exceeds  that  for  the  total 
Lummaton  sample  and  is  less  than  that  for  the  sample  from  nearby  Barton. 

There  is  usually  a  considerable  range  in  the  C.V.  values  for  colonies  of  the 
same  species  or  subspecies,  even  when  the  number  of  measured  corallites  in 
each  colony  is  the  same  or  nearly  so.  This  can  be  illustrated  with  reference  to 
Frechastraea  goldfussi  (Table  n)  in  which  the  C.V.  values  for  10  colonies  range 
from  io*oi  to  4*12,  the  value  for  the  total  sample  from  Ramsleigh  Quarry  being 
10 -io. 

As  coral  colonies  are  ideally  the  result  of  asexual  reproduction  from  one  sexually 
produced  individual,  variation  would  be  expected  to  be  lower  in  a  single  colony  than 
in  a  sample  of  several  colonies.  From  the  results  obtained  here  this  is  generally 
substantiated.  The  wide  range  in  colonial  variability  may  be  due  in  part  to  several 
factors.  Microenvironmental  and  general  ecological  influences  undoubtedly  exercise 
some  control  on  variation  but  their  effects  cannot  be  easily  assessed.  In  the  case 
of  phaceloid  colonies,  high  C.V.  values  may  be  the  result  of  the  intergrowth  of  two 
or  more  colonies  which  have  been  sampled  as  one.  As  Oliver  (i960  :  74)  pointed  out, 
it  is  virtually  impossible  to  detect  intergrowth  when  collecting  material,  and  this 
may  well  have  been  responsible  for  the  high  C.V.  values  in  the  colonies  of  Thamno- 
phyllum germanicum  schouppei  (Table  13).  Another  factor  for  consideration  is  the 
fusion  of  several  sexually  produced  polyps  at  an  early  stage  of  colony  formation, 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

-a 

be  mnwinoviionvd     oo  >o        ooo        n  u-o       ir>  oo  ms 

(3  '  ,  iO   iO  t^oo  O^OrOt^  VO    N 

as  utnsoiiqsdvo 

J-i 


c/0    g 
3  CL 


5  e 

y  .-Si 


ft, 


in  Oi        ro  o        oo    m        co  oo 


oo    h         wt^        wm         o    Tf  r^  oo 

SlSU3llA31fVS  N     H  in  00  W    IN  00     CO  lO   ID 


"d" 


p44no^os 


in  o        m  o        Oi  ■*       n  n 


ON  00  Ol^NMOOl-l  Tf    CO 

1    "  umotuvuuaS         M  °  ?  ^       V  *       ?  T1  *   ^ 

rt   n'  '  M-i-  N*         IDN  fON  ON 

en   "1  J^  m  h  m 

f  1  C«a) 

w>     ?S  .11  Tf   N  Oi    M  N   lO  O     Tf 


M/»MM^  ^    ^  ^    ^  ^  o 


fl      O  .„T  M  H  M  N 


to 

J$"  e  (•uruirri) 


g  vuoSvfiiaif 

^  vuo£v}U9<f 

'd 


N     CO  f^    0">  N    lO 


C  ^  iqvuuaq  ro  w       vo  oo 

c3    [[  iijvuu9q  n   o        in   is  io  oo 

§   §  uaqssn 


Tf  H 

O  r-~ 

vO  C~- 

lO  N 

0  <o 

Tf   O 

M   0> 

O   H 

rt  -g  Tf 

X3^>  Tf  ro          Tf  On                             Tfoo                             if)  (O 

o  a,  lijuvqAannoq                        in  i^       >o  >o                     n   Tf                     oh 

"3     ||  'if                                             Ht^CiN                              O    »0                             OvO 

el 


tssnfpioS  u-i  m 

•►r  HO 


o>  00 
O  CO 

N  lO 

M   M 

VO   Tf 

co  co 

ON 

t^  co 

o  NO 

M   Q\ 

0   CI") 

on  NO 

O  oo 

Oiu  CO 

o'n  n  no        uo  h  no   f^  o  oo 

rrt.  ^  VfVUUVD  o>or^io  <oo  oo 

§    CJO  'J"  h    0  On   CO  On   t^  Ot^ 

BO    -S  M 

pi    ?2 

g    Jj  VtUlUlW  NO  00  rf  N  «">0  no    h 

«  .9  vuoSv}us4  ??f>^  ?r  ?  ? 

>HH  .     ,-  OOONTf  OO  ON 

43  o  a  m  M  H 

u  h 
PhTTJ 
So  >>>>>>> 

o  .3  ix  d     ix  d     ix  d     ix  d     ix  d     ix  d     ix  d 

IN 

w 

^  S         ■«         ^ 

H  •o-oeiflC'd^ 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  203 

recorded  in  Recent  corals  by  Stephenson  (1931  :  124).  The  effect  of  this  phenomenon 
would  generally  be  to  increase  the  variation  displayed  by  the  colony. 

It  is  very  difficult  to  separate  the  effects  due  to  these  different  factors  but  it 
seems  likely  from  the  wide  range  in  C.V.  values  shown,  for  example,  by  the  colonies 
of  F.  goldfussi,  that  primary  polyp  fusion  could  be  an  important  factor. 

The  investigation  of  colonial  variation  described  here  differs  from  that  made  by 
Oliver  (i960  :  73  et  seq.)  through  the  latter  comparing  septal  ratio  C.V.  values  for 
individual  colonies  with  those  for  populations  of  solitary  cylindrical  corals.  In 
addition,  Oliver  analysed  the  total  sample  of  Tryplasma  fascicularia  from  colony 
means  rather  than  the  basic  corallite  data. 

(iii)  Variation  in  septal  number  and  the  septal  ratios.  With  the  exception  of 
Thamnophyllum  germanicum  schouppei,  the  C.V.  values  for  septal  number  are  very 
similar  in  all  the  total  samples.  There  appears  to  be  no  relationship  between  the 
degree  of  variation  and  the  mean  septal  number  (Table  2). 

It  has  long  been  known  that  septal  number  depends  to  some  extent  on  calice  size. 
When  one  is  plotted  against  the  other  for  both  colonies  and  total  samples,  n  shows 
a  general  increase  with  increasing  d  or  dt  in  every  case.  Correlation  coefficients  in 
the  total  samples  are  usually  about  0*9.  The  figure  in  individual  colonies  is  slightly 
lower  but  much  the  same  from  colony  to  colony.  Only  rarely  does  the  correlation 
coefficient  drop  below  07.  These  figures  reflect  the  strength  of  the  linear  relation- 
ship between  n  and  d  or  dt  in  mature  individuals.  A  correlation  between  septal 
number  and  diameter  is  still  reflected  to  a  large  extent  in  Table  2  as  taxa  with  larger 
mean  diameters  have,  with  few  exceptions,  greater  mean  values  for  n. 

This  relationship  had  led  to  the  use  of  the  septal  ratio  (septal  coefficient  of 
Voynovskiy-Kriger  1954)  as  a  useful  diagnostic  criterion  in  coral  species.  Rozkowska 
(1957)  particularly,  has  used  a  form  of  this  ratio,  her  Ms,  in  a  detailed  statistical 
study  of  species  of  Thamnophyllum  and  Macgeea  from  Poland.  She  showed  (p.  91 
and  Table  6)  that  successively  younger  species  and  subspecies  of  Thamnophyllum 
have  lower  Ms  values  and  similarly,  fewer  septa  at  a  given  diameter.  A  com- 
parison of  Rozkowska's  figures  with  the  thamnophyllids  described  here  is  shown  in 
Table  3. 

It  can  be  seen  that  the  n  values  at  d  =  6  mm.  for  the  English  material  follow  the 
same  trend  as  for  the  Polish  specimens  but  do  not  fit  exactly  into  the  latter's  scale. 
On  the  other  hand,  the  English  Ms  values  (the  English  data  has  been  recalculated 
for  direct  comparison  with  Rozkowska's  figures)  are  quite  different  from  their 
approximate  age  equivalents  in  Poland  and  do  not  fit  into  a  stratigraphical  trend. 
It  is  interesting,  however,  to  arrange  the  same  taxa  in  order  to  increasing  mean 
corallite  diameter.  This  produces  a  series  of  decreasing  Ms  values  somewhat  less 
perfect  than  before  for  the  Polish  specimens  but  into  which  the  English  figures  fit 
quite  well.  The  values  of  n,  however,  show  somewhat  less  perfect  ordering  than 
when  arranged  in  stratigraphical  series. 

Turning  to  the  massive  corals  in  Table  2,  the  mean  n/dt  figures  for  the  representa- 
tives of  Frechastraea  and  Phillipsastrea  show  a  perfect  series  of  decreasing  values 
with  increasing  mean  tabularium  diameter.  Furthermore,  just  such  correlation 
exists,  this  time  between  mean  d  and  mean  n/d,  in  the  corals  described  by  Oliver 


204 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


Table  3. — Series  in  Ms  and  n  (at  d  =  6  mm.)  for  English  (E)  and  Polish  (P)  thamnophyllids. 
Polish  figures  from  Rozkowska  (1957  :  91  and  table  6). 


d 

s 

w 

2 


Horizon  Name 

C  Middle        T.  super ius 


Country    Ms       n  (at 

d  =  6  mm.) 

P        2-20         15 


Lower         T.kozlowskii  P        2-30         17 

Upper         T '.  caespitosum  paucitabulatum    E        3-  50         i9'35 


T.  caespitosum  caespitosum 

P 

3-17 

20 

(T.  germanicum  pajchelae) 

P 

(3-6o) 

(21) 

Middle 

T.  caespitosum 

E 

3-15 

20 

T.  germanicum  germanicum 

P 

3*42 

21 

T.  germanicum  skalense 

P 

3-64 

22 

T.  germanicum  schouppei 

E 

4- 16 

20-28 

xd 

4-n 

T.  germanicum  pajchelae 

P 

3  60 

21 

4-14 

T.  germanicum  schouppei 

E 

4-16 

20-28 

5-36 

T.  germanicum  skalense 

P 

3-64 

22 

5-68 

T.  caespitosum  paucitabulatum 

E 

3-50 

19-35 

6- 00 

T.  caespitosum  caespitosum 

P 

3-17 

20 

6-17 

T.  caespitosum 

E 

315 

20 

7-00 

T.  germanicum  germanicum 

P 

3-42 

21 

9-28 

T.  superius 

P 

2-  20 

15 

10-03 

T.  kozlowskii 

P 

2-30 

17 

(ca.) 


(i960).  Thus  the  correlation  between  mean  size  and  mean  septal  ratio  is  quite 
strong.  The  stratigraphical  series  in  Ms  values  obtained  by  Rozkowska  may  simply 
reflect  the  tendency  in  her  material  for  larger  species  and  subspecies  of  Thamno- 
phyllum  to  occur  at  higher  horizons. 

Between  conspecific  colonies,  the  septal  or  septal-tabularium  ratio  behaves  in  the 
same  way.  Colonies  with  greater  mean  diameters  have,  with  very  few  exceptions, 
smaller  values  for  the  mean  septal  ratio. 

This  trend  is  related  to  the  fact  that  the  septal  ratio  is  not  constant  throughout 
ontogeny  but  decreases  in  value  (see  Voynovskiy-Kriger  1954)  with  increasing 
corallite  diameter.  When  septal  ratio  is  plotted  against  diameter,  the  relationship 
is  more  or  less  linear  for  mature  corallites.  In  the  present  case  correlation  coefficients 
for  the  total  samples  range  between  —  0-90  and  —  0*94  for  phaceloid  and  —  0-96 
and  —  i-o  for  massive  corals. 

G.V.  values  for  septal  number  and  the  septal  or  septal-tabularium  ratio  in  con- 
specific  colonies  may  show  a  considerable  range.  Furthermore,  the  relative  variation 
in  these  characters  and  d  or  dt  between  colonies  is  usually  different,  although  there 
is  a  slight  tendency  for  the  colony  with  the  highest  C.V.  value  for  diameter  to  have 
high  G.V.  values  for  septal  number  and  the  septal  ratio  as  well  (see  for  example 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  205 

Table  11).  In  the  case  of  n,  the  C.V.  value  reflects  to  some  extent  the  rate  of  septal 
insertion  as  well  as  the  amount  of  variation  in  diameter,  and  this  masks  any  inde- 
pendent variation  in  the  septal  number.  Colonies  with  similar  variation  in  diameter, 
for  example,  show  a  strong  correlation  between  a,  the  growth  ratio,  in  graphs  plotting 
n  against  d  or  dt  and  C.V.  values  for  n.  Similarly,  C.V.  values  for  n/d  or  n/dt  in 
conspecific  colonies  also  correlate  roughly  with  values  of  a  for  graphs  plotting  n/d 
or  n/dt  against  d  or  dt,  and  inversely  with  values  of  a  for  the  plots  of  n  against 
d  or  dt. 

(iv)  Variation  in  size  ratios.  In  the  phaceloid  forms,  tabularium  diameter 
increases  evenly  with  increase  in  corallite  diameter.  The  relationship  between  the 
two  is  strongly  linear,  reflected  in  the  high  values  for  the  correlation  coefficient 
(r  >  o*95  in  all  cases).  Only  three  phaceloid  colonies  have  been  analysed  individually, 
those  of  Thamnophyllum  germanicum  schoup-pei.  Their  correlation  lines  on  the  graph 
plotting  dt  against  d  are  virtually  superimposed  (Text-fig.  20). 

The  ratios  of  tabularium  to  corallite  diameter  have  very  low  C.V.  values,  partially 
reflecting  the  fact  that  they  are  virtually  unaffected  by  changes  in  diameter. 
Measurements  are  confined,  however,  to  mature  corallites.  Oliver  (i960  :  71)  has 
shown  that  during  ontogeny  in  Siphonophrentis  variabilis  and  Pseudoblothrophyllum 
helderbergium,  this  ratio  decreases  with  increasing  diameter.  In  his  analyses  of 
mature  individuals,  on  the  other  hand,  the  ratio  remains  relatively  constant  with 
increasing  diameter,  as  is  shown  in  the  present  material. 

The  relationship  between  tabularium  and  corallite  size  in  massive  corals  is  con- 
sidered in  terms  of  their  respective  areas  and  is  restricted  to  total  samples.  Analyses 
of  A,  At,  and  the  At/A  ratio  show  these  characters  to  have  high  C.V.  values.  In 
the  same  sample  there  may  be  a  considerable  difference  between  the  C.V.  value  for 
A  and  that  for  At,  for  example  in  the  case  of  Phillipsastrea  hennahi  ussheri  in  which 
the  former  figure  is  more  than  double  the  latter  (Table  5).  This  is  undoubtedly 
due  in  part  to  the  difficulty  in  calculating  accurately  the  value  of  A.  In  most  cases, 
however,  the  two  figures  are  more  nearly  comparable. 

When  At  is  plotted  against  A,  there  is  always  a  clear  tendency  for  the  former  to 
increase  with  increase  in  the  latter  (Text-figs.  12,  19).  The  scatter  of  points,  how- 
ever, is  considerable  and  is  reflected  in  the  low  values  for  the  correlation  coefficient. 
Calculations  using  the  formulae  for  the  fitted  lines  show  that  the  At/A  ratio  may 
increase  or  decrease  slightly  with  increasing  corallite  area.  Thus  the  analyses  suggest 
that  this  ratio  behaves  similarly,  if  not  so  regularly,  as  the  dt/d  ratio  and  is  largely 
independent  of  size  in  mature  colonies. 

V.  SYSTEMATIC  DESCRIPTIONS 

The  family  name  Phillipsastraeidae  was  erected  by  Roemer  (1883  :  389)  to  include 
the  genera  Phillipsastrea  and  Pachyphyllum.  It  has,  however,  been  little  used  until 
recently  and  taxonomists  have  usually  classified  the  nominal  type  genus  in  the 
Disphyllidae.  Hill  (1939  :  224)  erected  the  Disphyllidae  to  include  both  the  typical 
Disphyllum  and  the  related  Prismatophyllum  (considered  by  Lang,  Smith  &  Thomas 
1940  :  104  as  a  junior  synonym  of  Hexagonaria)  as  well  as  Phillipsastrea  (inter- 
preted by  Hill  1939  :  236  to  contain  species  both   with  and  without  horseshoe 


206  COLONIAL  PHILLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

dissepiments)  and  the  exclusively  horseshoe  bearing  genera,  Thamnophyllum,  Macgeea 
and  Trapezophyllum.  Hill's  concept  of  the  Disphyllidae  was  largely  influenced  by 
Lang  &  Smith's  (1935)  discussion  of  these  genera. 

Many  subsequent  workers  followed  Hill's  combination  of  horseshoe  and  non- 
horseshoe  bearing  forms  in  the  same  family.  Stumm  (1949  :  31)  extended  the  range 
of  genera  included  in  the  Disphyllidae  but  divided  them  among  three  subfamilies, 
the  Pachyphyllinae,  Disphyllinae  and  Eridophyllinae,  characterised  respectively  by 
the  presence  and  absence  of  horseshoe  dissepiments  and  the  development  of  an 
aulos.  He  considered  Phillipsastrea  to  lack  horseshoe  dissepiments  and  placed  it 
with  Disphyllum,  Hexagonaria  and  Billingsastraea  in  the  Disphyllinae.  The  Pachy- 
phyllinae included  Pachyphyllum,  Macgeea,  Thamnophyllum  and  Trapezophyllum. 

Soshkina  (1949),  on  the  other  hand,  distributed  the  genera  and  species  here 
included  in  the  Phillipsastraeidae  among  three  new  families  on  the  basis  of  morpho- 
logical and  ontogenetic  considerations.  Genera  she  considered  to  develop  horseshoe 
dissepiments  and  a  "  hexacoralloid "  microstructure — Pachyphyllum,  Macgeea, 
Thamnophyllum  and  Synaptophyllum — Soshkina  (1949  :  76)  grouped  in  the  Thamno- 
phylhdae,  equivalent  to  Stumm's  Pachyphyllinae.  She  placed  Phillipsastrea, 
erroneously  quoting  (1951  :  95)  P.  radiata  as  type  species  and  interpreting  the  genus 
as  lacking  horseshoe  dissepiments,  with  Neocolumnaria  and  Schluteria  in  the  Neo- 
columnariidae  (1949  :  145).  The  third  new  family,  the  Peneckiellidae  (1949  :  141) 
included  Peneckiella  and  Megaphyllum.  The  latter  was  considered  by  Hill 
(1956  :  280)  to  be  a  synonym  of  Disphyllum. 

Soshkina  (1951, 1952, 1954)  continued  to  use  this  classification  with  the  introduction 
of  further  genera  to  the  latter  two  families  and  (1954  :  44)  replacing  the  name 
Neocolumnariidae  by  Neocampophyllidae.  Spassky  (i960)  followed  Soshkina's 
(1954)  classification. 

Wang  (1950  :  217)  further  enlarged  the  Disphyllidae  by  incorporating  the  acantho- 
phyllids  into  the  family.  He  based  his  classification  on  a  consideration  of  coral 
microstructure,  defining  the  family  chiefly  on  the  development  of  one  or  more  fan 
systems  in  the  septal  trabeculae.  Wang's  use  of  the  subfamily  PhacellophyUinae 
(septal  trabeculae  with  a  marked  area  of  divergence)  corresponds  very  closely  to  the 
scope  of  the  Phillipsastraeidae  as  interpreted  herein.  Like  Lang  &  Smith  (1935) 
and  Hill  (1939),  however,  he  included  species  both  with  and  without  horseshoe 
dissepiments  in  Phillipsastrea. 

Rozkowska  (1953  :  8  et  seq.)  considered  the  three  different  classifications  of  Stumm 
(1949),  Wang  (1950)  and  Soshkina  (1951).  She  decided  that  coral  microstructure 
was  of  particular  diagnostic  importance  and  for  this  reason  followed  Wang's  classi- 
fication in  principle.  She  did,  however,  remove  the  Acanthophyllinae  from  the 
Disphyllidae.  In  addition,  she  used  the  subfamily  PachphyUinae  sensu  Stumm  in 
preference  to  the  PhacellophyUinae  sensu  Wang  as  the  former  was  conceived  as 
exclusive  of  the  genus  Phillipsastrea.  This  genus  Rozkowska  interpreted  as  lacking 
horseshoe  dissepiments  and  placed  in  the  subfamily  Disphyllinae. 

Hill  (1954a,  b)  reintroduced  the  family  name  Phillipsastreidae  (sic)  effectively  as 
a  senior  synonym  of  her  (1939  :  224)  Disphyllidae.  Although  the  1954  papers 
contain  no  familial  diagnosis,  this  classification  was  later  given  in  full  by  Hill 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  207 

(1956  :  179).  She  considered  Phillipsastrea  to  lack  horseshoe  dissepiments  and  thus 
placed  all  the  genera  with  disphyllid  and  marisastrid  dissepimentaria  in  the  Phillip- 
sastraeinae,  and  the  horseshoe  bearing  genera  in  the  Phacellophyllinae. 

Schouppe  (1956)  discussed  at  length  the  classifications  of  previous  authors.  He 
stressed  the  importance  of  the  so-called  "  hexacoralloid  "  microstructure  developed 
in  these  corals  with  strongly  reflexed  dissepimentaria,  and  advocated  their  clear 
systematic  separation.  Two  years  later,  Schouppe  (1958)  published  a  classification 
on  this  basis,  stressing  at  the  same  time  the  presence  of  horseshoe  dissepiments  in 
the  lectotype  of  Phillipsastrea  hennahi.  He  placed  all  the  genera  with  an  area  of 
divergence  in  their  septal  trabeculae,  usually  but  not  always  associated  with  horseshoe 
dissepiments,  in  the  Phillipsastraeacea  (sic).  Thus  he  elevated  what  had  previously 
been  a  family  or  even  a  subfamily  concept  to  the  level  of  a  suborder. 

Schouppe  subdivided  the  Phillipsastraeacea  into  the  Macgeeidae,  with  subfamilies 
Macgeeinae  and  Peneckiellinae,  and  the  Phillipsastraeidae.  He  placed  the  genera 
Phillipsastrea  and  Billingsastraea  together  in  the  Phillipsastraeidae  and  listed 
(1958  :  233)  the  family  characteristics  as  massive  astraeoid  form,  with  a  pseudotheca 
and  never  an  epitheca  between  adjacent  corallites,  and  with  a  broad  dissepimentarium 
often  developing  horseshoe  dissepiments.  Phillipsastrea  and  Billingsastraea,  how- 
ever, are  not  considered  to  be  closely  related  (Oliver  1964  :  2;  Scrutton  1967  :  276). 
Furthermore,  Schouppe  placed  Haplothecia  Freeh,  Pachyphyllum  sensu  Rozkowska 
(1953)  and  Pseudoacervularia  sensu  Rozkowska  (1953)  in  his  synonymy  for  Phillip- 
sastrea, all  of  which  include  species  with  a  partial  or  complete  epitheca  around  some 
or  all  corallites.  The  Macgeeidae  sensu  Schouppe,  on  the  other  hand,  was  defined 
by  the  presence  of  horseshoe  and  usually  also  flat  dissepiments  in  a  narrow  dissepi- 
mentarium, with  an  epitheca  surrounding  individual  corallites.  All  the  genera 
included  in  this  family  by  Schouppe,  with  the  exception  of  Synaptophyllum,  belong 
to  the  Phillipsastraeidae  as  defined  herein. 

Rozkowska  (1957  :  82)  rejected  her  earlier  classification  in  favour  of  that  proposed 
by  Soshkina  (1949)  to  the  extent  of  placing  all  horseshoe  dissepimentate  genera  in 
the  Thamnophyllidae.  Rozkowska  referred  to  Schouppe's  (1956  :  151)  views  but 
decided  to  separate  those  forms  with  horseshoe  dissepiments  from  those  without 
among  the  group  with  trabecular  fans.  Rozkowska  (1965  :  261),  however,  accepted 
almost  completely  Schouppe's  (1958)  classification  and  erected  a  new  family,  the 
Marisastridae,  for  Phillipsastraeacea  with  an  epitheca  but  no  horseshoe  dissepiments 
(see  Scrutton  1967). 

On  the  other  hand,  Strusz  (1965)  rejected  Schouppe's  (1958)  familial  and  ordinal 
groupings,  placing  all  genera  with  horseshoe  dissepiments  in  the  Phacellophyllidae. 
Genera  with  trabecular  fans  but  no  horseshoe  dissepiments  he  included  in  the 
Disphyllidae.  Strusz  (1965  :  523)  drew  attention  to  a  distinction  between  forms 
with  "  half  fans  "  and  full  "  disphylloid  fans  "  within  this  family  but  placed  no 
particular  taxonomic  significance  upon  it. 

The  confused  classification  of  this  group  of  corals  reflects  the  difficulty  of 
distinguishing  clear  phylogenetic  relationships  between  the  genera  involved.  Mor- 
phology is  extremely  variable,  particularly  in  massive  forms,  and  results  in  general 
gradations  between  the  characters  considered  diagnostic  of  different  family  and  sub- 


208  COLONIAL  PHI  LLI  PS  ASTR  AEI  D  AE  FROM  S.E.  DEVON 

family  groups.  A  particular  feature  of  more  recent  classifications  resulting  from  this 
is  the  uncertainty  in  placing  corals  lacking  horseshoe  dissepiments  but  having  a  fan 
shaped  arrangement  of  the  septal  trabeculae.  Schouppe  (1958)  grouped  them  with 
the  horseshoe  bearing  forms  whilst  Strusz  (1965)  grouped  them  with  the  disphyllids 
sensu  stricto.  All  workers  are  agreed,  however,  that  horseshoe  dissepiments  and  the 
related  trabecular  fans  must  be  considered  of  particular  importance  in  taxonomy. 
Horseshoe  dissepiments  are  unique  to  this  group  of  Devonian  corals  and  great  stress 
has  been  laid,  particularly  by  Soshkina  (1949),  Schouppe  (1956)  and  Rozkowska 
(1957)  on  the  development  of  a  pseudohexacoralloid  trabecular  pattern. 

The  trabecular  arrangement  alone,  however,  has,  in  the  writer's  opinion,  been 
somewhat  overstressed  and  does  not  justify  the  rank  of  suborder  sensu  Schouppe 
(1958).  On  the  other  hand,  the  development  of  specialized  dissepimental  types 
(horseshoe  and  peneckielloid  dissepiments),  with  their  related  trabecular  structure, 
defines  with  relative  clarity  a  group  of  corals  whose  general  morphological  character- 
istics support  a  close  family  relationship.  This  group  includes  the  genus  Phillip- 
sastrea  (as  defined  herein)  and  should  be  classified  as  the  Phillipsastraeidae.  The 
writer  does  not  agree  with  Strusz  (1965  :  524)  concerning  the  use  of  this  family  name. 
It  must  be  noted  that  the  generic  name  Phillipsastrea  has  been  used  far  more  often 
in  a  disphyllid  sense  than  the  corresponding  family  name.  The  genus,  however, 
should  not  be  suppressed  simply  because  its  type  species  has  been  imperfectly  known 
in  the  past  and  in  this  case,  the  change  in  concept  of  the  family  name  will  naturally 
follow  that  of  the  genus. 

The  genera  with  reflexed  dissepimentaria  lacking  horseshoe  or  peneckielloid 
dissepiments  but  possessing  an  open  fan  shaped  arrangement  of  the  trabeculae  form 
a  group  of  their  own,  intermediate  in  character  between  the  Phillipsastraeidae  as 
defined  herein  and  the  disphyllids  sensu  stricto.  This  is  formalized  in  the  family 
Marisastridae  sensu  Scrutton  (1967).  Finally,  the  Disphyllidae  is  here  restricted 
to  forms  in  which  the  trabeculae  are  arranged  in  half  fans,  or  in  a  sub-parallel  sense 
throughout  on  non-reflexed  dissepimentaria.  This  family,  in  the  writer's  opinion, 
should  be  interpreted  to  conform  strictly  to  the  dissepimental  pattern  of  the 
type  genus  Disphyllum. 

Family  PHILLIPSASTRAEIDAE  Roemer  1883 

1883  Phillipsastraeidae  Roemer  389. 

partim  1922  Campophyllidae  Wedekind  3. 

partim  1929  Pexiphyllidae  Walther  117. 

partim  1939  Campophyllidae;  Soshkina  :  12. 

partim  1939  Cyathophyllidae ;  Sanford  :  408. 

partim  1939  Disphyllidae  Hill  224. 

partim  1940  Disphyllidae;  Hill  :  258. 

partim  1942c  Disphyllidae;  Hill  :  186. 

partim  1949  Disphyllidae;  Stumm  :  31. 

partim  1949  Thamnophyllidae  Soshkina  :  76. 

partim  1949  Peneckiellidae  Soshkina  :  141. 

partim  1949  Neocolumnariidae  Soshkina  :  145. 

partim  1950  Disphyllidae;  Wang  :  217. 

partim  1951  Disphyllidae;  Taylor  :  183. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  209 

partim  1952  Disphyllidae ;  Lecompte  :  470. 

partim  1953  Disphyllidae;  Rozkowska  :  9. 

partim  1954a  Phillipsastreidae;  Hill  :  14. 

partim  1954b  Phillipsastreidae;  Hill  :  107. 

partim  1955  Columnariidae;  Glinski  :  86. 

partim  1956  Phillipsastraeidae ;  Hill  :  279. 

1957  Thamnophyllidae ;  Rozkowska  :  83. 

partim  1958  Macgeeidae;  Schoupp6  :  218. 

partim  1958  Phillipsastraeidae;  Schouppe  :  232. 

partim  1959  Disphyllidae;  Middleton  :  152. 

partim  1959  Phillipsastraeidae;  McLaren  :  22. 

partim  i960  Thamnophyllidae;  Spassky  :  44. 

partim  1961  Phillipsastraeidae;  Semenoff-Tian-Chansky  :  294. 

1962  Thamnophyllidae;  Soshkina  :  308. 

partim  1962  Disphyllidae;  Soshkina  &  Dobrolubova  :  334. 

partim  1962  Phillipsastraeidae;  Soshkina  &  Dobrolubova  :  336. 

1964  Phacellophyllidae ;  Pedder  :  366. 

1965  Phacellophyllidae;  Strusz  :  554. 

1966  Phacellophyllidae;  Pedder  :  183. 

Type  genus.    Phillipsastrea  d'Orbigny  1849  :  12. 

Diagnosis.  Solitary,  dendroid,  phaceloid  or  massive  rugose  corals.  Septa  of 
two  orders,  major  and  minor,  more  or  less  dilated  at  the  tabularium  boundary  which 
is  usually  sharply  defined.  Characterized  by  an  area  of  divergence  in  the  septal 
trabeculae  centred  on  a  complete  or  incomplete  series  of  horseshoe  dissepiments, 
peneckielloid  dissepiments  or  a  series  of  highly  globose  dissepiments  occasionally 
modified  to  a  horseshoe  form. 

Distribution.  Lower  and  particularly  Middle  and  Upper  Devonian  of  Europe, 
Asia,  Australia  and  North  America. 

Discussion.  Of  the  genera  assigned  to  this  family,  all  but  two  have  a  more  or 
less  well  developed  series  of  horseshoe  dissepiments.  The  exceptions  are  Peneckiella 
and  Frechastraea  gen.  nov. 

Peneckiella  may  show  considerable  variation  in  dissepimental  form  but  is  charac- 
terized by  the  development  of  peneckielloid  dissepiments.  Frequently  these  are 
accompanied  by  sigmoidal,  horseshoe  and  flat  dissepiments  in  varying  proportions. 
Rozkowska  (i960  :  32,  48,  50)  who  named  the  Peneckiella  dissepimental  types  con- 
sidered them  to  be  the  breakdown  products  of  a  typical  Thamnophyllum  dissepi- 
mentarium,  and  that  Peneckiella  evolved  from  this  genus. 

The  species  and  subspecies  of  Frechastraea  develop  rare  horseshoe  dissepiments 
against  the  tabularium  junction.  This  genus  is  thought  to  have  evolved  from 
Phillipsastrea  principally  through  the  gradual  loss  of  horseshoe  dissepiments  and 
thus  should  be  classified  with  the  Phillipsastraeidae. 

The  full  list  of  genera  here  included  in  the  Phillipsastraeidae  is  as  follows : 
Phillipsastrea  d'Orbigny   1849   (Synonyms  Smithia,  Pachyphyllum,   Medusae- 

phyllum,  IStreptastrea,  Pseudoacervularia  and  ? Keriophylloides). 
Macgeea  Webster  1889  (Synonyms  Pexiphyllum  and  Protomacgeea). 
Thamnophyllum  Penecke  1894  (Synonym  Phacellophyllum). 
Trapezophyllum  Etheridge  1899. 


212  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

1849  Actinocyathus  d'Orbigny  :  12. 

1850  Syvingophyllum  Edwards  &  Haime  :  72. 

1 85 1  Phillipsastrea;  Edwards  &  Haime  :  173,  447. 

1852  Phillipsastraea;  Edwards  &  Haime  :  203. 
1855  Syvingophyllum;  Edwards  &  Haime  :  295. 

1951  Phillipsastraea;  Soshkina  :  95. 

1952  Phillipsastraea;  Soshkina  :  101. 

1953  Phillipsastraea;  Rozkowska  :  57. 

1954  Phillipsastraea;  Soshkina  :  45. 
1958  Phillipsastraea;  Bulvanker  :  118. 
i960  Phillipsastraea;  Spassky  :  65. 

Diagnosis.  Cerioid,  pseudocerioid,  astraeoid,  thamnasterioid,  aphroid  or  rarely 
secondarily  phaceloid,  rugose  corals.  Major  and  minor  septa  with  spindle-shaped 
dilatation  at  tabularium  boundary.  In  dissepimentarium,  a  series  of  horseshoe 
dissepiments  is  well  developed  at  or  near  junction  with  tabularium;  several  series 
of  normal  dissepiments  also  present.     Tabulae  complete  or  incomplete. 

Type  species.  (Selected  by  Edwards  &  Haime  1850  :  71).  Astrea  hennahii 
Lonsdale  (1840  :  697,  pi.  58,  figs.  3,  36,  non  fig.  3a)  =  Astraea  hennahii  Lonsdale; 
Phillips  (1841  :  12,  pi.  6,  figs.  i6aa,  i6/?b,  i6/?c,  non  pi.  7,  fig.  15D) :  upper  Givetian 
limestones;   Barton  Quarry,  Torquay. 

Distribution.  Middle  and  Upper  Devonian  of  Europe  and  Asia.  Lower  and 
Middle  Devonian  of  Australia.     Upper  Devonian  of  North  America. 

Discussion.  Phillipsastrea  was  erected  by  d'Orbigny  (1849  :  12)  who  quoted  as 
examples  of  the  genus  "  Astrea  parallela  et  hennahii  Phillips  ".  In  the  following 
year  he  redefined  the  genus  (1850  :  107),  spelling  it  this  time  Phillipsastraea  and 
splitting  up  the  specimens  illustrated  by  Lonsdale  (1840)  and  Phillips  (1841)  as 
A.  hennahi  among  the  genera  Phillipsastrea,  Lithostrotion  and  Actinocyathus. 
D'Orbigny's  original  spelling  Phillipsastrea  is  employed  here  following  Hill  (1956  : 
279)  who  considered  the  spelling  Phillipsastraea  to  be  a  nomen  vanum. 

The  confusion  created  by  Edwards  &  Haime  (1850  :  70;  185 1  :  173)  who  quoted 
first  "  Astrea  hennahi  Lonsdale  "  and  later  Erismatolithus  Madreporites  radiatus 
Martin  (1809,  pi.  18)  as  type  species  of  Phillipsastrea  has  been  clearly  documented 
by  Smith  (1917  :  284).  He  showed  the  selection  of  A.  hennahi  to  be  valid  and  placed 
Phillipsastrea  radiata  and  other  Carboniferous  species  assigned  to  Phillipsastrea  in 
the  genera  Aulina  and  Orionastraea. 

Edwards  &  Haime's  invalid  designation  however  was  still  accepted  by  Russian 
workers  up  to  the  beginning  of  this  decade  (see  Soshkina  1954  :  45;  Bulvanker 
1958  :  118;  Spassky  i960  :  65)  and  Soshkina  (1952  :  86)  described  A.  hennahi 
Lonsdale  as  a  species  of  Pachyphyllum.  More  recently,  however,  Soshkina  & 
Dobrolubova  (1962  :  336)  have  assigned  the  correct  type  species  to  Phillipsastrea. 

Various  authors  in  the  last  century,  particularly  Edwards  &  Haime,  assigned  a 
number  of  Devonian  colonial  corals  to  Acervularia  Schweigger.  In  fact,  Edwards  & 
Haime  (1850  :  70)  unwarrantably  cite  Acervularia  roemeri  de  Verneuil  &  Haime  as 
type  species  of  that  genus.  Acervularia,  however,  the  type  species,  by  monotypy, 
of  which  is  Madrepora  ananas  Linnaeus  1758  :  797  =  Acervularia  baltica  Schweigger 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  213 

1819,  Table  6;  Upper  Silurian,  Gotland,  Sweden,  is  typically  Silurian  with  only  a 
superficial  resemblance  to  the  Devonian  forms.  With  the  closer  study  of  internal 
structures  in  later  years,  the  Devonian  species  of  Acervularia  were  removed  from 
that  genus  and  usually  reassigned  to  Phillipsastrea  (see  Freeh  1885  :  44).  Schliiter 
(1881  :  84)  had,  in  fact,  erected  a  new  genus  Pseudoacervularia  expressly  to  cover 
the  Devonian  "  acervulariids  ",  but  this  genus  has  been  little  used.  Indeed,  a  type 
species  was  not  selected  until  Lang,  Smith  &  Thomas  (1940  :  108)  chose  Acervularia 
coronata  Edwards  &  Haime  for  that  purpose.  Unfortunately  the  type  specimens  of 
A.  coronata  are  missing  but  the  species  is  almost  certainly  conspecific  with  Phillip- 
sastrea hennahi,  and  Pseudoacervularia  is  a  subjective  synonym  of  Phillipsastrea. 

Rozkowska  (1953  :  39),  apparently  unaware  of  Lang,  Smith  &  Thomas'  selection, 
invahdly  chose  Acervularia  macrommata  F.  A.  Roemer  as  type  species  of  Pseudo- 
acervularia. She  described  several  species  of  that  genus,  all  of  which  should  be 
placed  in  Phillipsastrea. 

A  most  critical  factor  in  the  taxonomic  position  and  interpretation  of  Phillipsastrea 
has  been  the  emphasis  placed  by  Schouppe  (1958  :  156)  on  the  development  of 
horseshoe  dissepiments  in  the  lectotype  of  the  type  species.  Hitherto,  the  presence 
of  these  specialized  dissepiments  in  the  lectotype  had  not  been  widely  realized  and 
the  genus  had  been  either  restricted  to  species  lacking  horseshoe  dissepiments 
(Stumm  1949  :  34;  Rozkowska  1953  :  57;  Hill  1956  :  280),  or  considered  to  include 
species  both  with  and  without  horseshoe  dissepiments  (Lang  &  Smith  1935  :  556; 
Hill  1939  :  236;  Smith  1945  :  36).  In  addition,  Edwards  &  Haime  (1850  :  68) 
had  erected  the  genus  Pachyphyllum  with  P.  bouchardi  as  type  species.  The  latter 
has  long  been  known  to  develop  horseshoe  dissepiments  and  many  authors  (Stumm 
1949  :  37;  Soshkina  1951  :  84;  Rozkowska  1953  :  39)  have  used  the  genus  specific- 
ally for  Middle  and  Upper  Devonian  massive  philhpsastreids  with  well  developed 
horseshoe  dissepiments. 

Schouppe  (1958  :  233  et  seq.),  however,  placed  Pachyphyllum  in  synonymy  with 
Phillipsastrea.  At  the  same  time,  he  removed  all  the  forms  previously  assigned 
to  the  latter  genus  which  he  thought  to  lack  horseshoe  dissepiments  and  placed  them 
in  Billingsastraea  Grabau  (see  comments  under  that  genus  by  Scrutton  (1967  :  276) 
and  under  Frechastraea  gen.  nov.  herein). 

The  writer  agrees  with  Schouppe's  concept  of  Phillipsastrea  with  the  exception  of 
the  placing  of  Haplothecia  Freeh.  Schouppe  (1958  :  201)  claimed  that  the  type  species 
of  the  latter  developed  horseshoe  dissepiments.  However,  an  examination  of  the 
lectotype  of  H.  filata  (Schlotheim),  type  species  of  Haplothecia,  does  not  substantiate 
this  (Scrutton  1967  :  271)  and  the  genus  is  here  regarded  as  quite  distinct  from 
Phillipsastrea. 

Semenoff-Tian-Chansky  (1961  :  303)  was  not  convinced  that  Phillipsastrea  and 
Pachyphyllum  should  be  considered  congeneric  and  he  mentioned  two  particular 
points  of  apparent  difference  in  overall  size  and  dissepimental  arrangement.  With 
regard  to  the  considerable  disparity  in  size  between  the  respective  type  species,  the 
existence  of  species  of  intermediate  size  must  not  be  overlooked — for  example 
Phillipsastrea  ananas  (Goldfuss),  P.  devoniensis  (Edwards  &  Haime),  P.  ibergensis 
(F.  A.   Roemer)   and  P.   chenouensis   (Semenoff-Tian-Chansky).     In  fact  it  would 

geol.  15,  5.  23 


214  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

appear  that,  quite  fortuitously,  Phillipsastrea  hennahi  is  one  of  the  smallest  and 
Pachyphyllum  bouchardi  the  largest  of  the  species  assigned  to  Phillipsastrea.  The 
apparent  discrepancy  between  Pachyphyllum  and  Phillipsastrea  in  the  arrangement 
of  the  normal  dissepiments  was  based  on  a  comparison  with  Schouppe's  (1958,  text- 
fig.  21;  pi.  5,  fig.  1)  longitudinal  illustrations  of  P.  hennahi.  As  remarked  elsewhere, 
although  purporting  to  come  from  the  lectotype,  they  are  completely  different  from 
longitudinal  slides,  and  peels  taken  from  that  specimen  by  the  writer,  which  are 
illustrated  here  (PI.  1,  figs.  2-4,  6).  There  is  in  fact  no  basic  difference  in  the  dissepi- 
mental  arrangement  between  P.  hennahi  and  P.  bouchardi. 

Phillipsastrea  is  thus  interpreted  as  including  all  those  Devonian  massive  corals, 
excepting  species  of  Trapezophyllum,  Sulcorphyllum  and  Bensonastraea,  in  which  a 
series  of  horseshoe  dissepiments  is  well  developed  in  the  dissepimentarium.  Trapezo- 
phyllum Etheridge  (1899  :  32)  differs  from  Phillipsastrea  in  possessing  only  one  row 
of  flat  dissepiments  peripheral  to  the  horseshoe  series  and  in  this  respect  is  related 
more  closely  to  Thamnophyllum.  Sulcorphyllum  Pedder  (1964  :  366)  is  like  Trapezo- 
phyllum but  with  several  rows  of  normal  dissepiments  separating  the  peripheral  flat 
dissepiments  from  the  horseshoes  adjacent  to  the  tabularium.  The  irregular, 
incomplete  tabulae  of  this  genus  are  unlike  those  found  in  European  phillipsastreids. 
Sulcorphyllum  is,  for  the  present,  considered  distinct  from  Phillipsastrea  but  a  better 
knowledge  of  the  variation  in  the  former  may  require  its  taxonomic  position  to  be 
reconsidered  in  the  future.  Bensonastraea  Pedder  (1966  :  183)  is  distinguished  by 
its  vepreculate  and  peripherally  degenerate  septa,  and  complex  dissepimentarium. 

Phillipsastrea  hennahi  hennahi  (Lonsdale) 
Plate  1,  figs.  1-6;  Plate  2,  figs.  1-4 

1840  Astrea  (Siderastrea)  hennahii  Lonsdale  :  697  pars,  pi.  58,  figs.  3,  36  (non  3a). 

1841  Astraea  hennahii  Lonsdale;  Phillips  :  12  pars,  pi.  6,  figs.  i6aa,  i6^3b,  i6/3c  (non  pi.  7, 
fig.  15D). 

1841  Astraea  inter cellulosa  Phillips  :  12,  pi.  6,  fig.  17. 

?  1843  Astraea  hennahii  Lonsdale;   F.  A.  Roemer  :  5,  pi.  2,  fig.  13. 

1849  Phillipsastrea  hennahii  (Phillips)  d'Orbigny  :  12  pars. 

1850  Lithostrotion  hennahii  (Lonsdale)  d'Orbigny  :  106. 
1850  Favastraea  inter  cellulosa  (Phillips)  d'Orbigny  :  107. 
1850  Actinocyathus  hennahii  (Phillips)  d'Orbigny  :  107. 

1850  Phillipsastraea  hennahii  (Phillips);   d'Orbigny  :  10 7  pars. 

?  1850  Acervularia  roemeri  de  Verneuil  &  Haime  :  162. 

1850  Phillipsastrea  cantabrica  de  Verneuil  &  Haime  :  162  (nomen  nudum). 

1850  Phillipsastrea  hennahi  (Lonsdale);   Edwards  &  Haime  :  71  pars. 

1851  Acervularia  coronata  Edwards  &  Haime  :  416. 

?  1851  Acervularia  roemeri  de  Verneuil  &  Haime;   Edwards  &  Haime  :  420. 

1 85 1  Smithia  hennahii  (Lonsdale)  Edwards  &  Haime  :  421. 

1 85 1  Syringophyllum?  cantabricum  (de  Verneuil  &  Haime)  Edwards  &  Haime  :  451. 

1851  Arachnophyllum  hennahi  (Lonsdale)  M'Coy:    72. 

1853  Acervularia  coronata  Edwards  &  Haime;    Edwards  &  Haime:    237,  pi.  53,  figs.  4a-b 

(•'fig-  4)- 
1853     Acervularia  roemeri  de  Verneuil  &  Haime;    Edwards  &  Haime  :  239,  pi.  54,  figs.  3,  3a. 
1853     Smithia  hennahii  (Lonsdale) ;   Edwards  &  Haime  :  240,  pi.  54,  figs.  4-4^. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  215 

1853     Syringophyllum  cantabricum  (de  Verneuil  &  Haime) ;    Edwards  &  Haime  :  242,  pi.  55, 

figs.  3~3«- 

?  1855  Smithia  hennahi  (Phillips);   F.  A.  Roemer  :  33,  pi.  6,  fig.  25. 

?  1856  Streptastrea  longiradiata  Sandberger  &  Sandberger  :  416,  pi.  37,  figs.  3-36. 

?  1879  Acervularia  roemeri  de  Verneuil  &  Haime;   Quenstedt  :  535,  pi.  162,  fig.  38. 

1879  Astrea  hennahii  Lonsdale;    Quenstedt  :  535. 

1883  Acervularia  coronata  Edwards  &  Haime;   C.  F.  Roemer  :  352. 

?  1883  Acervularia  roemeri  de  Verneuil  &  Haime;   C.  F.  Roemer  :  353. 

1883  Phillipsastraea  hennahii  (Lonsdale);  C.  F.  Roemer  :  390. 

1885  Phillipsastrea   roemeri    (de   Verneuil    &    Haime)    Freeh  :  57   pars,    pi.    4,  ?  fig.  2  (non 
figs.  1,  3-5). 

1885  Phillipsastrea  hennahi  (Lonsdale);   Freeh  :  59  pars,  pi.  5,  fig.  1,  [non  figs.  2-4). 

1917  Phillipsastraea  hennahi  (Lonsdale);    Smith  :  284,  pi.  22,  figs.  1-4. 

1945  Phillipsastraea  hennahi  (Lonsdale);    Smith  :  37,  pi.  19,  figs.  la,  b. 

1952  Pachyphyllum  hennahi  (Lonsdale)  Soshkina  :  86. 

1958  Phillipsastraea  hennahi  (Lonsdale);  Schouppe  :  235,  ?text-figs.  20,  21,  ?pl.  5,  fig.  1. 

1959  Phillipsastraea  hennahi  (Lonsdale);   Middleton  :  156. 

1963     Phillipsastraea  hennahi  (Lonsdale) ;  Fontaine  :  84,  pi.  8,  figs.  5,  6. 

But  not: 

1953  Pseudoacervularia  roemeri  (de  Verneuil  &  Haime)  R6zkowska:  53,  text-figs.  27,  28,  pi. 
7.  figs-  3.  4- 

Diagnosis.  Astraeoid  tending  to  thamnasterioid  Phillipsastrea.  Mean  tabu- 
larium  diameter  2*45  mm.  with  10  to  16  major  septa  (topotype  sample).  Septa 
thin  peripherally,  with  a  variable  spindle-shaped  thickening  at  tabularium  junction. 
Dissepimentarium  composed  of  several  rows  of  normal  dissepiments  with  a  single 
series  of  horseshoe  dissepiments  developed  against  tabularium.  Tabulae  complete 
or  incomplete,  usually  in  form  of  wide  flat  plates  with  downturned  edges.  Increase 
lateral. 

Lectotype.  Selected  by  Edwards  &  Haime  (1851  :  421).  The  original  of 
Lonsdale's  (1840,  pi.  58)  figures  3  and  36  which  is  GSM  (Geol.  Soc.  Coll.)  6185;  upper 
Givetian  limestones;  Barton  Quarry,  Torquay. 

Material.  Barton  Quarry:  OUM  D514  (Colony  1),  OUM  D306  (Colony  2), 
OUM  D512  (Colony  3),  BM(NH)  R5615  (Colony  4).  Other  measured  specimens: 
GSM  (Geol.  Soc.  CoU.)  6185  (lectotype),  OUM  D515-6,  OUM  D518-20,  OUM  D522, 
TM(JB)  107-8,  TM(JB)  120.  Additional  material:  OUM  D240  (=  D282),  OUM 
D513,  OUM  D517,  OUM  D523,  TM(JB)  139. 

Lummaton  Quarry:  TM(JB)  61  (Colony  1),  TM(JB)  59  (Colony  2),  BM(NH) 
R23484  (Colony  3).  Other  measured  specimens:  TM(JB)  63-4,  TM(JB)  78-9. 
Additional  material:  TM(JB)  62,  TM(JB)  69. 

Wolborough  Quarry :   OUM  D521. 

Babbacombe  Cliff :  BM(NH)  R46176. 

Dartington  Hall:   GVMi3/i,  GVM14/2. 

Torquay,  South  Devon:  OUM  D74  (holotype,  Astraea  inter cellulosa  Phillips), 
GSM  (Geol.  Soc.  Coll.)  6189  (lectotype  here  chosen,  Syringophyllum  cantabricum 
Edwards  &  Haime). 


216  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

Distribution.  England:  upper  Givetian  limestones  of  Barton  and  Lummaton 
quarries,  Torquay;  middle  Givetian  limestones  of  Wolborough  Quarry,  Newton 
Abbot;  Givetian  limestones,  ioo  ft.  above  sea  level,  south  end  Babbacombe  cliffs, 
Torquay;  middle  Givetian  limestones  of  Dartington  Hall  area  (Middleton  1959); 
Middle  Devonian,  Plymouth  (Phillips  1841).  Devonian,  Kikai,  Yunnan  (Fontaine 
1963).  Also  represented  in  the  Frasnian,  Ibergerkalk,  Bad  Grund,  Harz,  Germany 
(Freeh  1885).  Middle  Devonian,  Pola  de  Gordon,  Leon,  Spain  (de  Verneuil 
&  Haime  1850). 

Description.  All  specimens  are  incomplete  and  details  of  the  colony  exterior, 
its  overall  size  and  shape  are  unknown.  The  largest  specimen  examined  was,  when 
complete,  17  X  14  cm.  in  surface  area  and  5  cm.  deep  (it  is  now  in  two  pieces  numbered 
separately  as  OUM  D240  and  OUM  D282). 

Colonies  are  astraeoid  tending  to  thamnasterioid,  with  a  straight  or  zigzagged 
pseudotheca  of  variable  strength  separating  the  corallites.  This  pseudotheca  may 
break  down  round  parts  of  the  corallite  margins  when  the  septa  run  more  or  less 
confluently  from  one  corallite  to  the  next. 

The  septa,  major  and  minor,  are  very  thin  peripherally,  less  than  o*i  mm.  and 
normally  about  0*025  mm.  in  thickness.  They  may  have  smooth  or  slightly 
roughened  sides  and  rarely  some  separation  of  the  trabeculae  but  they  are  never 
truly  carinate.  They  are  usually  sinuous,  seldom  straight,  in  the  dissepimentarium 
and  occasionally  curved  in  a  constant  direction  to  form  a  vortex.  The  septa  develop 
a  spindle-shaped  dilatation  in  the  region  of  the  tabularium  junction.  It  is  extremely 
variable  both  within  colonies  and  individual  corallites,  ranging  normally  between 
0-15  and  0*25  mm.  width  and  1  to  2  mm.  length  (see,  for  example  TM(JB)79,  PI.  2, 
fig.  I). 

Within  the  tabularium,  the  major  septa  usually  taper  smoothly  and  end  somewhat 
short  of  the  axis  of  the  corallite.  Occasionally,  however,  some  may  reach  the  axis, 
or  cross  the  tabularium  periaxially  to  join  with  major  septa  in  the  adjacent  quadrant 
of  the  corallite.  Rarely  the  axial  ends  of  the  major  septa  may  develop  slight  lobate 
thickenings.     The  minor  septa  do  not  enter  the  tabularium. 

Dissepiments  are  usually  uniserial  but  may  occasionally  be  multiserial  between 
adjacent  septa.  In  cross-section  they  are  well  spaced  peripherally,  becoming  more 
crowded  towards  the  tabularium  junction  which  is  sharply  defined.  Against  this 
junction,  the  trace  of  a  single  series  of  horseshoe  dissepiments  is  sometimes  clearly 
seen  (PI.  2,  fig.  2  corallite  in  right  centre). 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  rows  of 
small,  normal  dissepiments,  between  0*2  and  0-5  mm.  in  height.  They  are  weakly 
arched  peripherally  but  become  increasingly  globose  towards  the  tabularium 
boundary.  The  dissepiments  immediately  adjacent  to  the  tabularium  are  modified 
into  horseshoe  form  resulting  in  a  single  vertical  series  of  more  or  less  well  developed 
horseshoe  dissepiments  (PI.  I,  figs.  2-6).  The  surface  of  the  dissepimentarium  slopes 
downwards  away  from  this  series  and  flattens  out  peripherally. 

The  tabulae  are  complete  or  incomplete,  more  usually  the  latter.  In  general  they 
are  strikingly  flat  and  parallel  across  the  axis  of  the  tabularium,  usually  with  down- 
turned  peripheral  edges.     There  is  a  limited  development  of  vesicular,  periaxial 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 
90 


80 


217 


70 


60 


50 


40 


30 


20 


10 


0  l 


145 


15 
14 
1  3 
12 
1  1 
10 


s 

/ 

/ 

s' 

s 

/ 

/^ 

-^"^        /s' 

/'    ,'' 

/    ,' 

/     ' 

1  85  2-25  2-65  305 

Ot      (  0-2  m  m    class     intervals) 


7  r 


3-45 


/at 


16  20  24  2-8  32  36 

dt  (mm) 


16  20  2-4  2-8 

dt  (mm) 


3-2  3-6 


Total    sample  Colony 


Colony     2 


Colony      3 


C  olony      4 


Fig.  9.     Phillipsastvea  hennahi  hennahi  (Barton  sample). 


218 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


c6 


in 


d 
o 
"o 
o 


d 
o 
"o 

o 


*-~. 

>, 

o 

d 

-A'a  ro 

o 

*  S 

"o 

ft  s 

o 

6  <° 

rt    d 

en    0 

d    rt 

%    6 

—  Jn 

S  8 
ffl  3 

13  "cL 
o  E 

ti 

H  <* 

k       w 

a  § 

►S      ■» 
^    >« 

1"^ 

s  « 

S  s 

s  a 

Tj- 

>N 

d 

a  <a 

o 

oi  ^ 

■&  -& 

"o 

co     to 

O 

e   « 

CO       05 

.•^.■g. 

S     3 

co 

►a  -a 

>> 

0,0, 

d 
o 

M-t  <+h 

o  o 

"o 

._    tn 

0 

d 
o 

"o 
Q 


o 

>, 

N+H 

d 

JO 

13 

U 

H 

o 

■*H 

■4-1 

^ 

tn 

+3 

o  S 

CO 

en 

M 

a 

1 


-t- 
o> 


00 

in 

O 

t^ 

ro 

r| 

IT) 

t^ 

Th 

On 

in  no 

r^ 

a\ 

tN 

CT> 

M 

on 

On  NO 

m 

c 

o 

•* 

H 

o 

tH 

A 

CO 

in 

-t- 

0 

NO 

o 

H 

CO 

o 

00 

Tt- 

Tt- 

Tf 

On 

in 

tN 

b 

IO 

0 

CI 

0 

~t 

0 

oo 

«■> 

O 

<N 

o 

6 

>*■ 

M 

O 

6 

t-^ 

M 

H 

in 

1 

1 

ro 

00 

O   i-~  in  n 

if   N  H     MO 

t(-  m    On   O 

H    ir)  +  N    O 

tN     O    NO     O 


m    tN    0    ro  O 


t 


< 

O     tN 
•      rt- 
O    0 

o 
o 

w     O 
ro  O 

On   ro 
tN    O 

On  NO   NO 

00 
OO 

t^  NO 

< 

4o 

ro 
O 

0 

b 

^  O 

in   O 

O     N   NO 

6 

1 

H      On 
1 

f-~ 

o 

H 

NO 

tN 

co 

On 

On 

ro 

O 

tt- 

o 

Tt" 

o 

H 

ti-oo 

tN 

<*■ 

-I- 

H 

-t- 

t(- 

O 

0 

On  no 

o 

CT- 

Th 

in  io 

Tf 

+J 

o 

o 

0 

o 

o 

N 

CO 

rN 

V) 

O 

M 

ro 

<*■ 

CO 

O 

i 

-t- 

in 

N 

O 

t^  NO 

Tt- 

On 

O 

l-l 

< 

1 

o 

0 

~t- 

o 

1 
H 

<N 

O 

o 

M 

o 

o 

H 

H 

0 

ro 

0 

I 

ro 
tN 

in 

0 

On 

o 

o 

M 

00 

O 
1 

tN 
1 

0 

ro 
O 

H 

0 

•     N    tJ-  ro 

ro  ro  tN    h 

>n 

M 

0 

in  t^ 

tH     NO 

O    N 

M      M 

t^    ON 
CO  00 

ro 
tN 
0 

00 

r^                 oo 
•     n    in  no    in 
t^  in  ^  w    o 

t^   tN     O     O 

o 

O     CN 

o 

ro  m  O  oo    o 

CN 

M 

M- 

ro  oo 
I      • 

H      N 


in  o    t^  i-i 

H     tfiON 


O  oo    o 


o 

Ti-    O 


t^.     Th    Tt"    <N 


ro   ■*■  O    t(-  O 


O  00 

•  in  w    On   ro 

CO  in   CN     O     O 

H  tN     O   00     O 

rN 


Oifl    ONO    O 


rN  w  m  n 
I  •  •  ■ 
On    tS     O     t> 


00 

in  ro  t^  m 
rf  ro  M    O 


On 
CO  ro  co  h    r^ 

H    O    Th  no    O 

w    CN    O    CO  O 


ir,NN    H 


m  oo 

7  ?  . 

N    CO  o 


in 

O    w  oo 

NO     NO      O 


00     O 

O    w 

Tf  b 


tN    O    ro  O 


ti-  tJ-  O  no    O 


no    tN    N    ro  O 

tN  >n  o   in  o 


o 
o 

•    00 

t^    IN 

I    • 

tN    NO 

Tf- 
in 


ro  ro 

in    '-i  On  ro 

N   N    NO  lO  O 

On  in  o  m  o 


oo  no    t^ 

oo   in  o 


a 

.   s 

a 

rt         >  w 

«        >a 

^         >.  w 

i 

0  IX    85  U  CC 

0  IX  «)  U  03 

0  ix  »  U  t» 

J. 

d 

•o 

tn 

-d 
ft 
ctj 

l-i 

O 

!-     S    .2 


OO  N    NO 

On  On    On 

O  m    On 

I  I 


Ph 
ctj 
u 

o 


Tt-NO 
Ol    N     H 

vo  o  o 


h    tSJ3 


ro 

H 
| 

H 

o 

m 

-i- 

in 

o 
o 

NO 

w 

b 

tN 

0 

b 

. 

S 

«. 

>  w 

0  IX  ce  U  CO 


O 

in 

0 

M 

CO 

ON 

TT- 

in 

ICOO 

^ 

■* 

H 

H 

O 

1>N 

"0 

t-t 

l>.NO 

ro 

o 

o 

0 

■* 

0 

o 

ro 

•+ 

o 
1 

1 

On 

>n 

rO 
O 

0 

b 

ro 

0 

ro    ro   O 

00    NO     ON 


O      H 
I  I 


P_ 
crj 
u 

o 


m  co 
oo  in  h 
t^  o   o 


^   d  £ 


■>*■ 
so  o  O  m 
o   cn   t^  o 

H    o  oo    o 


e 

0  IX  oo  O  85 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


219 


no 


100 


90 


80 


70 


60 


50 


40      • 


30     ■ 


20 


10 


1  45 


185  2  25  2  65 

dt      (0  2  mm   class   intervals) 


305 


14 


13 


n     12    • 


11     ■ 


10 


16  20  2-4  28 


dt 


/dt 


20 
dt 


2-4  28 


Total   sample 


Colony       1 


Colony       2 


Colony       3 


Fig.  10.     Phillipsastvea  hennahi  hennahi  (Lummaton  sample) 


220  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

elements.  Tabularium  structure  varies  in  complexity  from  the  simple,  interleaved 
tabulae  of  TM(JB)  79  (PI.  i,  fig.  5)  to  the  close  spaced  series  of  flat-topped  domes  in 
GSM  (Geol.  Soc.  Coll.)  6185  (PI.  i,  figs.  2-4,  6). 

Despite  the  large  number  of  specimens  examined,  only  one  instance  of  lateral 
increase  has  been  seen  and  that  is  in  the  lectotype. 

Statistical  analyses  have  been  made  of  samples  of  this  subspecies  from  the  type 
locality,  Barton  Quarry  (14  colonies)  and  Lummaton  Quarry  (7  colonies),  both  in 
Torquay.  Four  colonies  from  Barton  and  three  from  Lummaton  have  been  analysed 
individually.  The  statistics  are  given  in  Table  4  and  illustrated  graphically  in 
Text-figs.  9,  10  and  12. 

The  topotypes  are  characterized  by  a  mean  tabularium  diameter  of  2^45  mm.  with 
10  to  16  major  septa  and  a  mean  septal-tabularium  ratio  of  5*25.  The  Lummaton 
specimens  differ  principally  in  having  a  smaller  overall  range  and  mean  dt,  and  a 
somewhat  larger  mean  n/dt. 

Colonial  variation  at  the  two  localities  displays  an  interesting  contrast.  Generally 
there  is  a  greater  range  in  the  mean  values  of  the  quantitative  characters  between 
the  colonies  from  Barton,  although  variation  within  each  colony  is  much  the  same. 
The  Lummaton  colonies,  on  the  other  hand,  have  very  similar  mean  values  but 
markedly  different  degrees  of  variation  from  colony  to  colony. 

Discussion.  The  specimen  on  which  Phillips  (1841  :  12)  erected  Astraea  inter- 
cellulosa  is  OUM  D74  (PI.  2,  figs.  2,  3).  Phillips  himself  had  great  doubts  as  to 
whether  or  not  his  specimen  was  distinct  from  Lonsdale's  species.  In  the  writer's 
opinion  it  is  an  atypical  representative  of  P.  hennahi  hennahi  in  which  the  septa  are 
virtually  unthickened. 

F.  A.  Roemer  (1843  :  5,  pi.  2,  fig.  13)  described  and  figured  as  Astraea  hennahi 
Lonsdale  a  specimen  which  was  later  selected  by  de  Verneuil  &  Haime  (1850  :  162) 
as  the  type  of  a  new  species  Acervularia  roemeri.  The  identity  of  Roemer's  original 
is  uncertain  but  Edwards  &  Haime  (1853  :  239,  pi.  54,  figs.  3,  3a)  later  recorded  A. 
roemeri  from  Torquay.  The  latter's  specimen,  judging  from  their  figures,  is  almost 
certainly  consubspecific  with  P.  hennahi  hennahi.  Freeh  (1885,  pi.  4,  figs.  1-5)  also 
figured  Phillipsastrea  roemeri  (de  Verneuil  &  Haime),  of  which,  if  his  interpretation 
was  correct,  the  original  of  his  fig.  2  is  a  topotype.  Again  it  appears  from  the 
illustration  that  the  specimen  may  be  consubspecific  with  Phillipsastrea  hennahi 
hennahi. 

Edwards  &  Haime  (1853)  described  and  figured  two  species,  Acervularia  coronata 
and  Syringophyllum  cantabricum,  previously  erected  by  them  in  1851  (p.  416  and 
p.  451  respectively).  In  the  case  of  A.  coronata,  Edwards  &  Haime's  original  material 
has  been  lost  or  mislaid.  From  their  illustrations  (1853,  pi-  53.  ngs-  4a>  °)  OI  a 
specimen  from  Barton,  however,  the  species  is  almost  certainly  the  same  as  the 
present  subspecies.  "Phillipsastrea  cantabrica  n.  sp.  "  in  de  Verneuil  &  Haime 
(1850  :  162)  is  a  nomen  nudum  as  they  give  neither  description  nor  figures  of  the  species. 
Edward  &  Haime  (1851  :  451)  must  therefore  be  regarded  as  the  authors  of  the 
species  with  their  description  of  "  Syringophyllum?  cantabricum  (de  Verneuil  & 
Haime)  ".  The  original  of  Edwards  &  Haime's  (1853,  pi.  55,  figs.  3,  3a)  figures  of 
Syringophyllum  cantabricum  is  specimen  GSM  (Geol.  Soc.  Coll.)  6189,  which  is  here 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  221 

chosen  as  lectotype  of  the  species  (PL  2,  fig.  4).     This  specimen  is  consubspecific 
with  P.  hennahi  hennahi. 

Schouppe  (1958,  text-fig.  21,  pi.  5,  fig.  1)  published  two  illustrations  which  he 
stated  were  taken  from  a  longitudinal  peel  of  specimen  GSM  (Geol.  Soc.  Coll.)  6185. 
The  structure  of  both  the  tabularium  and  the  dissepimentarium  in  Schouppe's 
figures,  however,  is  not  that  seen  in  this  specimen  although  the  presence  of  horseshoe 
dissepiments  can  be  confirmed.  The  source  of  Schouppe's  figures  must  thus  be 
considered  somewhat  enigmatic. 

Phillipsastrea  hennahi  ussheri  subsp.  nov. 

Plate  3,  figs.  1-3 

Derivation  of  name.  The  subspecies  is  named  after  W.  A.  E.  Ussher  (1849- 
1920). 

Diagnosis.  Astraeoid  tending  to  thamnasterioid  Phillipsastrea.  Mean  tabular- 
ium diameter  2-04  mm.  with  10  to  14  major  septa  (topotype  sample).  Septa  thin 
peripherally  but  usually  strongly  dilated  at  tabularium  junction,  becoming  laterally 
contiguous  and  forming  a  dense  inner  wall.  Horseshoe  dissepiments  developed  in 
a  series  adjacent  to  tabularium  with  several  rows  of  normal  dissepiments  peripherally. 
Tabulae  usually  incomplete  with  narrow  axial  series  of  flat  topped  domes  irregularly 
developed.     Increase  lateral. 

Holotype.  OUM  D544.  Lower  Frasnian  limestones ;  road  cutting,  20  yd.  west 
of  the  entrance  to  Ramsleigh  Quarry,  East  Ogwell,  near  Newton  Abbot. 

Material.  Ramsleigh  Quarry:  BM(NH)  R23402  (Colony  i),  BM(NH)  R5616 
(Colony  2),  BM(NH)  R23209  (Colony  3). 

Road  cutting,  40  yd.  west  of  Ramsleigh  Ouarry  entrance.  Measured  specimens : 
OUM  D533-4- 

Road  cutting,  20  yd.  west  of  Ramsleigh  Quarry  entrance.     OUM  D545. 

Distribution.     Lower  Frasnian  limestones  in  and  around  Ramsleigh  Quarry. 

Description.  Only  incomplete  colonies  have  been  found  and  the  overall  colony 
shape,  size  and  external  features  are  unknown. 

The  colonies  are  astraeoid  tending  to  thamnasterioid  with  considerable  variation 
in  the  strength  of  the  pseudotheca.  Perfectly  confluent  septa  between  adjacent 
corallites,  however,  are  rare.  Septa  are  thin  peripherally,  usually  about  0-06  mm. 
thick,  occasionally  straight  but  more  usually  sinuous.  The  sides  of  the  septa  may 
be  smooth  or  slightly  crenulate;  carinae  are  not  developed.  Immediately  adjacent 
to  the  tabularium,  the  septa  are  dilated.  The  short,  spindle-shaped  thickening  is 
greater  on  the  major  septa  when  it  is  usually  0-5  to  07  mm.  long  and  about  0*25  mm. 
wide.  The  dilatation  is  a  very  constant  feature,  frequently  bringing  adjacent 
septa  into  contact  to  form  a  dense  wall,  about  0-5  mm.  wide,  round  the  tabularium. 
Minor  septa  end  at  the  tabularium  junction  but  the  major  septa  thin  abruptly  and 
continue  as  very  fine  filaments,  about  0-02  mm.  thick,  towards  the  axis  of  the  corallite. 
Sometimes  the  ends  of  the  major  septa,  slightly  withdrawn  from  the  axis,  bear  small 
lobate  thickenings  which  through  lateral  contact  may  form  a  pseudoaulos.     In 


224  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

strong  concentric  walls  (PI.  3,  fig.  1),  or  delimit  the  width  of  the  single  thick  wall 
when  the  septa  are  in  lateral  contact  (PL  3,  fig.  3). 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  rows  of 
small,  globose  dissepiments,  usually  between  o*i  and  0*3  mm.  in  height.  Immedi- 
ately adjacent  to  the  tabularium,  the  dissepiments  are  quite  sharply  modified  into 
a  vertical  series  of  horseshoes,  predominantly  uniserial  and  quite  regularly  developed. 
The  surface  of  the  dissepimentarium  slopes  steeply  away  from  the  series  of  horseshoe 
dissepiments  but  is  flat  over  most  of  its  area.  The  tabularium  junction  is  sharply 
defined. 

Tabularium  structure  is  not  perfectly  clear  from  the  present  material.  The 
tabulae  appear  to  be  closely  spaced  and  usually  incomplete.  Often  there  is  a  wide 
peripheral  series  of  horizontal  plates,  slightly  distally  concave,  abutting  against  a 
central  series  of  small,  flat-topped  or  occasionally  globular  domes,  which  occupy  a 
third  to  a  quarter  of  the  tabularium  diameter.  In  other  portions  of  the  tabularium, 
this  axial  structure  is  missing  and  the  whole  width  is  occupied  by  flat,  interleaved 
plates. 

Four  examples  of  lateral  increase  have  been  seen  in  BM  (NH)  R5616  although  in 
each  case  the  new  individuals  are  well  advanced. 

A  small  sample  of  five  colonies  of  this  subspecies  has  been  statistically  analysed. 
Three  of  the  colonies  have  also  been  treated  separately  to  investigate  colonial 
variation.  The  resulting  statistics  are  given  in  Table  5  and  the  data  illustrated 
graphically  in  Text-figs.  11  and  12. 

Discussion.  It  is  possible  that  some  of  the  material  from  Upper  Devonian 
localities  in  continental  Europe  which  has  been  described  as  Phillipsastrea  hennahi 
properly  belongs  to  this  subspecies;  for  example,  the  specimen  figured  by  Freeh 
(1885,  pi.  5,  fig.  2)  from  the  Frasnian  Ibergerkalk  at  Bad  Grund,  Germany. 

Comparison  of  the  subspecies  of  Phillipsastrea  hennahi.  The  main  quantitative 
characteristics  of  P.  hennahi  hennahi  (Barton  Quarry  sample)  and  P.  hennahi  ussheri 
have  been  statistically  compared  and  the  results  are  tabulated  below : 

dt  t  =  14-69  (5240  freedom)  P  <^  o-oi 

n/dt  z  (slope)  =    7-38  P<^o-oi 

z  (slope)  =1-91  P  >  o-oi 

z  (position)  =    7-02  P<^o-oi 


At/A 


The  mean  tabularium  diameter  for  P.  hennahi  ussheri  (2*04  mm.)  is  significantly 
smaller  than  that  for  P.  hennahi  hennahi  (2-45  mm.).  There  is  also  a  significant 
difference  between  the  slopes  of  the  correlation  lines  on  the  graph  plotting  n  against 
dt  for  the  two  subspecies.  The  number  of  major  septa  increases  slightly  faster  with 
increasing  tabularium  diameter  in  the  Frasnian  subspecies.  The  number  of  points 
on  the  graph  plotting  At  against  A  is  rather  low  to  allow  the  results  of  a  "  z  "  test 
to  be  uncritically  accepted  (Imbrie  1956  :  237,  footnote  1).  Nevertheless,  the  signi- 
ficant difference  in  position  reflects  the  small  degree  of  overlap  of  the  points  for  the 
two  subspecies. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


225 


100 


80 


60 


40 


20     • 


0    l 


1  05 


A 

/A 


//  \ 

///  \        \ 


V.      \ 


145  185  2  25  2-65  3  05 

at        (0  2  mm    class   intervals) 


3  45 


07 


•06 


05 


At 

(sq.  cms) 


04 


■03 


•02 


S*  *'  ~"*' 


*s 


/ 


/ 


/     i  S.~ ?'  o • 


0  4 


06 


08  10 

A        (sq.  cms) 


12 


16 


15 
14 
13 
12 
11 
10 


% 


\ 


\ 


^ 


^ 


•^ 


12  16  20         24  2-8         32         36 

dt      (mm) 


12  16  20         24  28  32         3-6 

at      (mm) 


Phillipsastrea      hennahi      hennahi 


Pml/ipsosfrea      hennahi      ussher 


(x)   Barton       •— —       (°)    Lummaton 

(*)    East  Ogwell 


Fig.    12.     Graphical  comparison  of  some  quantitative  characters  for  the  subspecies  of 

Phillipsastrea  hennahi. 


226  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

Qualitatively  the  two  subspecies  are  very  similar  and  obviously  closely  related. 
The  strong  cross-sectional  traces  of  horseshoe  dissepiments,  usually  infilled  by  close 
packed  septal  tissue,  which  surrounds  the  tabularium  in  P.  hennahi  ussheri  is, 
however,  very  distinctive.  The  horseshoe  traces  are  rarely  so  clearly  seen  in  P. 
hennahi  hennahi  and  excessive  septal  dilatation  is  similarly  uncommon.  The  major 
septa  within  the  tabularium  are  also  contrasted  in  the  two  subspecies.  The  abrupt 
attenuation  of  the  septa,  which  closely  approach  the  axis,  often  with  a  lobate  thicken- 
ing of  their  axial  ends,  is  characteristic  of  P.  hennahi  ussheri.  The  major  septa  of 
P.  hennahi  hennahi  are  usually  somewhat  withdrawn  from  the  axis  and  taper  away 
from  the  dilated  zone. 

In  longitudinal-section,  the  most  noticeable  difference  between  the  two  subspecies 
is  in  the  tabularium  structure.  The  narrow  axial  domes  and  wide,  flat  peripheral 
plates  of  P.  hennahi  ussheri  contrast  with  the  broad  based,  flat-topped  domes  of 
P.  hennahi  hennahi.  There  is,  however,  no  characteristic  difference  between  the 
dissepimentaria  of  the  two  subspecies. 


Phillipsastrea  devoniensis  (Edwards  &  Haime) 

Plate  4,  figs,  r-4 

1 85 1  Pachyphyllum  devoniense  Edwards  &  Haime  :  397. 

1853  Pachyphyllum  devoniense  Edwards  &  Haime;   Edwards  &  Haime  :  234,  pi.  52,  figs.  5,  5a. 

1883  Pachyphyllum  devoniense  Edwards  &  Haime;   C.  F.  Roemer  :  393,  text-fig.  93. 

1885  Phillipsastrea  [Pachyphyllum)  devoniense  (Edwards  &  Haime)  Freeh  :  67,  pi.  6,  figs.  2,  za. 

1952  Pachyphyllum.  devoniense  Edwards  &  Haime;    Soshkina  :  86. 

But  not: 

1958     Pachyphyllum  devoniense  Edwards  &  Haime;    Bulvanker  :  93,  pi.  45,  figs.  1a,  b. 

Diagnosis.  Astraeoid  tending  to  thamnasterioid  Phillipsastrea  with  poorly 
defined  corallite  margins.  Mean  tabularium  diameter  372  mm.  with  16  to  23 
major  septa  (South  Devon  sample).  Septa  generally  thin,  with  slight,  elongate 
dilatation  surrounding  tabularium.  In  longitudinal-section,  a  series  of  horseshoe 
dissepiments  is  developed  just  outside  tabularium  junction.  Tabulae  wide,  flat, 
complete  or  incomplete. 

Holotype.  The  original  of  Edwards  &  Haime's  (1853,  pi.  52)  fig.  5  and  5«; 
Devonian,  Torquay.     This  specimen  is  either  mislaid  or  lost. 

Material.  "Rocky  Valley".  Measured  specimens:  OUM  D277,  BM(NH) 
74489. 

PLummaton  Quarry.     Measured  specimen:   TM(JB)  105. 
Torquay,  S.  Devon:   BM(NH)  R1456. 

Distribution.  England:  in  situ  only  from  middle  Givetian  limestones,  Wol- 
borough  Quarry,  Newton  Abbot  and  Pfrom  upper  Givetian  limestones,  Lummaton 
Quarry,  Torquay.  Other  specimens  are  beach  pebbles,  several  being  known  from 
"  Rocky  Valley "  (?  =  Valley  of  Rocks,  Watcombe,  north  of  Torquay).  Also 
recorded  from  the  Ibergerkalk,  Bad  Grand,  Germany;  Frasnian. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  227 

Description.  Nothing  is  known  of  the  overall  size,  shape  and  external  features 
of  this  species. 

Colonies  are  astraeoid  tending  to  thamnasterioid.  The  septa  are  usually  out- 
wardly geniculate  and  irregularly  abutting.  Occasionally  they  may  be  perfectly 
confluent  from  one  coraUite  to  the  next  or  rarely  they  end  in  the  dissepimentarium 
with  a  free  peripheral  end.  The  margins  of  the  corallites  are  indistinctly  defined  in 
the  absence  of  a  well  formed  pseudotheca. 

The  septa,  major  and  minor,  are  o-i  mm.  or  less  in  thickness  in  the  dissepimentarium 
where  they  follow  a  straight  or  sinuous  course.  In  a  zone  surrounding  the  tabularium 
of  1  to  1  *5  mm.  width,  the  septa  have  a  slight,  elongate  dilatation,  usually  to  about 
0-2  or  0-3  mm.  thickness.  Minor  septa  project  just  within  the  tabularium.  The 
major  septa  continue  in  a  strongly  attenuate  form  more  or  less  to  the  axis,  where 
their  ends  may  be  free  and  infrequently  slightly  thickened,  or  confluent  with  septa 
in  the  adjacent  quadrant  of  the  tabularium.     Carinae  are  not  developed. 

Dissepiments  are  usually  uniserial  between  adjacent  septa.  Just  outside  the 
tabularium,  in  the  zone  of  septal  dilatation,  cross-sectional  traces  of  horseshoe 
dissepiments  are  often  clearly  visible.     The  tabularium  junction  is  clearly  defined. 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  series  of 
arched  dissepiments  usually  between  o#2  and  0*4  mm.  high.  Just  outside  the 
tabularium,  a  slightly  discontinuous  single  series  of  irregularly  sized  horseshoe 
dissepiments  is  developed.  From  the  crest  of  the  series,  the  dissepiments  slope 
steeply  downwards  towards  the  coraUite  margins  where  they  become  flat  lying. 
On  the  axial  side  of  the  horseshoes,  one  or  two  rows  of  normal  dissepiments  are 
usually  present. 

The  tabularium,  which  is  more  or  less  sharply  delimited  from  the  dissepimentarium, 
is  composed  of  wide,  flat  or  gently  undulating  plates,  usually  slightly  downturned 
peripherally.  The  tabulae  are  complete  or  incomplete  and  a  periaxial  series  of 
sloping  plates  is  intermittently  developed. 

Insufficient  material  is  available  for  a  statistical  study  of  variation  in  this  species. 
The  few  measurements  made  are  summarized  in  Table  6. 

Discussion.  The  present  material  agrees  very  well  with  Edwards  &  Haime's 
(1853  :  234,  pi.  52,  figs.  5,  5«)  figures  and  descriptions,  although  their  original  speci- 
men appears  to  have  had  somewhat  larger  tabularia.  Freeh's  (1885)  figures  of  a 
specimen  from  the  Ibergerkalk  (Harz,  Germany)  are  very  close  to  the  illustrations  of 
the  holotype  and  to  the  present  material,  although  the  horseshoe  dissepiments  in  the 
German  specimen  appear  stronger  and  more  regularly  developed. 

Phillipsastrea  devoniensis  differs  markedly  from  P.  hennahi  in  both  tabularium 
size  and  septal  number.  In  addition,  the  horseshoe  dissepiments  of  the  former  are 
somewhat  less  regular  in  size  and,  unlike  those  of  P.  hennahi,  are  separated  from  the 
tabularium  by  steep  sloping  normal  dissepiments.  The  thin  elongate  septal  dilata- 
tion in  P.  devoniensis  (PI.  4,  fig.  2)  also  contrasts  with  the  markedly  spindle-shaped 
septa  of  P.  hennahi  (PL  1,  fig.  1). 

P.  devoniensis  is  similar  in  many  respects  to  P.  bouchardi  and  P.  ibergensis.  Both 
the  latter  species,  however,  have  slightly  different  tabularium  structure  and  P. 
bouchardi  has  much  larger  tabularia  (dt  =  10-16  mm.  according  to  Semenoff-Tian- 


228  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

Chansky  1961  :  307)  than  P.  devoniensis.  P.  ibergensis  is'further  distinguished  by 
strong  spindle  shaped  septal  dilatation. 

The  specimen  figured  by  Bulvanker  (1958,  pi.  45,  figs,  la,  b)  as  P.  devoniense  has 
considerably  larger  tabularia  than  the  English  material.  It  is  closer  in  size  to  P. 
bouchardi  and  also  has  the  slightly  sagging  tabularium  structure  of  that  species  in 
contrast  to  the  flat  or  faintly  domed  tabulae  of  P.  devoniensis. 

P.  devoniensis  is  an  uncommon  species.  Very  few  specimens  have  been  seen  in 
English  museums  and  the  author  has  found  none  in  the  field.  It  appears  that 
Freeh's  (1885  :  67)  specimen  is  the  only  record  of  the  species  from  abroad. 


Phillipsastrea  ananas  (Goldfuss) 
Plate  5,  figs.  1-4 

1826     Cyathophyllum  ananas  Goldfuss  :  60  pars,  pi.  19,  fig.  46,  non  4a. 

1 85 1     Acervularia  troscheli  Edwards  &  Haime  :  416. 

1885     Phillipsastrea  ananas  (Goldfuss)  Freeh  :  49  pars,  pi.  2,  figs.  1,  2,  5,  5a,  $b,  pi.  3,  ?fig.  5; 

pi.  8,  ?fig.  9- 
1953     Pseudoacervularia  ananas  (Goldfuss)  Rozkowska  :  52,  text-figs.  27,  28,  pi.  7,  figs.  1,  2. 

But  not: 

1881     Heliophyllum  troscheli  (Edwards  &  Haime);   Schliiter  :  203,  pi.  4,  figs.  3,  4. 

Diagnosis.  Pseudocerioid  Phillipsastrea.  Mean  tabularium  diameter  3-54  mm. 
with  14-18  major  septa  (East  Ogwell  sample).  Septa  thin  peripherally,  with  slim, 
spindle-shaped  dilatation  near  tabularium  boundary.  Major  septa  usually  form 
weak  axial  vortex.  Dissepimentarium  reflexed,  with  imperfect  series  of  horseshoe 
dissepiments  at  crest  from  which  two  or  three  rows  of  normal  dissepiments  slope 
steeply  into  tabularium.  Axial  tabulae  flat-topped  domes  with  peripheral  horizontal 
concave  plates. 

Type  specimen.  The  original  specimen  figured  by  Goldfuss  (1826,  pi.  19,  fig.  4b) 
is  either  mislaid  or  lost.     Frasnian;   Namur,  Belgium. 

Material.  BM  (N.H.)  R46158-9.  Road  cutting,  south  side,  30-35  yd.  west  of 
Ramsleigh  Quarry  entrance,  East  Ogwell,  near  Newton  Abbot ;   Lower  Frasnian. 

Distribution.  England:  Lower  Frasnian;  East  Ogwell,  near  Newton  Abbot. 
Also  widespread  in  Frasnian  of  Namur,  Belgium;  Harz,  Stolberg  near  Aachen, 
Germany;   Kielce  region,  Poland;   Urals,  Kuznetsk  Basin,  U.S.S.R. 

Description.  The  material  consists  of  fragments  of  apparently  pseudocerioid 
coralla.     The  epitheca  is  predominantly  straight  but  may  be  zigzagged. 

The  septa,  both  major  and  minor,  are  spindle-shaped.  At  the  periphery  they 
are  about  0-05  mm.  across  but  the  major  septa  may  dilate  up  to  0*45  mm.  thick  near 
the  tabularium  junction.  The  maximum  length  of  septal  thickening  is  about  2  mm. 
The  septa  may  be  carinate.  BM  (N.H.)  R46159  (PI.  5,  fig.  3)  has  well  developed 
yardarm  carinae,  up  to  0-25  mm.  across  with  a  minimum  spacing  of  0*15  mm.,  on 
many  but  not  all  septa.  BM  (N.H.)  R46158  (PI.  5,  fig.  2)  on  the  other  hand  has  only 
rarely  developed,  very  faint,  carinae. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  229 

The  minor  septa  do  not  penetrate  the  tabularium  but  the  major  septa  continue, 
much  attenuated,  almost  to  the  axis  where  they  usually  form  a  weak  vortex.  There 
is  usually  an  axial  area  0*2-0 -4  mm.  in  diameter  which  is  free  of  septa. 

The  dissepiments  may  be  uni-  or  multiserial  between  septa.  The  tabularium 
junction  is  not  strongly  marked  but  coincides  approximately  with  the  axial  ends  of 
the  minor  septa. 

Table  6. — Quantitative  comparison  of  some  species  and  subspecies  of  Phillipsastrea. 


dt 

n 
O.R. 

n/dt 
x 

At/A 

r 
O.R. 

1 

X 

x 

P.  hennahi  hennahi 

1-7-3-5 

2-45 

10-16 

5-25 

0-045 

P.  hennahi  ussheri 

1-2-2-5 

2-04 

10-14 

6-i6 

0-063 

P.  devoniensis 

3-0-5-0 

3-72 

16-23 

5-04 

ca. 

o- 1 

P.  ananas 

30-4-0 

3-54 

14-18 

4-44 

ca. 

0-15 

P.  rozkowskae 

2-3-2-9 

2-64 

12-15 

5-08 

ca. 

O-I 

In  longitudinal-section,  the  dissepimentarium  is  of  variable  width,  and  the  dissepi- 
mental  surface  strongly  reflexed  with  the  crest  on  average  o-6  mm.  outside  the 
tabularium.  The  dissepiments  at  the  crest  are  usually  modified  to  horseshoe  shape, 
forming  a  slightly  discontinuous  vertical  series  of  horseshoe  dissepiments.  There 
are  several  rows  of  normal  dissepiments  towards  the  margins  of  the  corallites  and 
on  the  other  side  of  the  crest,  two  or  three  rows  of  dissepiments  slope  steeply  into  the 
tabularium. 

The  tabulae  are  complete  or  incomplete  flat-topped  domes,  occupying  about 
three-fifths  of  the  tabularium  diameter.  There  is  a  peripheral  series  of  horizontal, 
distally  concave  plates. 

The  two  known  specimens  from  East  Ogwell  have  slightly  different  quantitative 
data  which  is  averaged  in  Table  6.  Individually,  BM  (N.H.)  R46159  has  tabu- 
larium diameters  of  3  to  3-5  mm.  with  14  or  15  major  septa  whilst  BM  (N.H.)  R46158 
has  3*8  to  4*0  mm.  tabularium  diameters  with  15  to  18  major  septa. 

Discussion.  Although  the  two  specimens  described  here  differ  slightly  in  some 
respects,  they  both  appear  to  fall  within  the  range  of  variation  shown  by  P.  ananas 
in  areas  where  the  species  is  more  abundant.  On  the  other  hand,  they  also  show 
some  features  which  are  intermediate  in  character  with  P.  macrommata  (F.  A.  Roemer). 

Freeh  (1885  :  49)  considered  these  two  species  to  be  synonymous  whilst  Rozkowska 
(1953  :  49-52)  was  able  to  distinguish  between  them  in  Poland  through  the  lower 
septal-tabularium  ratio,  the  short,  strong  septal  dilatation  and  the  slightly  different 
dissepimental  arrangement  in  P.  macrommata.  The  English  material  agrees  in 
general  morphological  appearance  with  P.  ananas  of  Rozkowska.  On  the  other 
hand,  the  septal  insertion  in  both  the  present  specimens  follows  more  closely  the 
curve  for  P.  macrommata  given  by  Rozkowska  (1957,  text-fig.  26). 

In  view  of  the  limited  material,  the  two  specimens  are  placed  in  P.  ananas  on 
morphological  grounds.  They  may,  however,  need  re-evaluation  if  a  larger  sample 
ever  becomes  available  from  East  Ogwell. 

GEOL.  15,  5.  24 


230  COLONIAL  PHILLIPS ASTRAEIDAE  FROM  S.E.  DEVON 

Phillipsastrea  rozkowskae  sp.  nov. 
Plate  6,  figs.  1-4 

Derivation  of  name.  The  species  is  named  after  Professor  Maria  Rozkowska 
(Poznan,  Poland). 

Diagnosis.  Pseudocerioid  Phillipsastrea.  Tabularium  diameter  a^^-c^mm. 
with  12  to  15  major  septa  (topotype  sample).  Septa  of  two  orders,  major  with 
short,  club-shaped  dilatation  against  tabularium  boundary,  becoming  extremely 
attenuate  in  tabularium;  minor  septa  hardly  thickened.  Dissepiments  in  several 
rows  with  single  vertically  discontinuous  series  of  horseshoe  dissepiments.  Tabular- 
ium structure  simple,  with  incomplete  tabulae. 

Holotype.  BM(NH)  R46156.  Lower  Frasnian;  road  cutting,  south  side, 
25  yd.  west  of  Ramsleigh  Quarry  entrance,  East  Ogwell,  near  Newton  Abbot. 

Material.     BM(NH)  R46156-57.     Horizon  and  locality  as  for  holotype. 

Distribution.     Type  locality  only. 

Description.  Massive,  pseudocerioid  colonies  with  predominantly  pentagonal 
corallites  separated  by  straight  to  slightly  zigzagged  pseudothecae.  Only  two  frag- 
ments are  available  and  details  of  colony  shape,  size  and  external  features  are 
unknown. 

The  septa,  major  and  minor,  are  0-05-0-15  mm.  thick  peripherally.  They  are 
generally  straight  but  may  occasionally  be  very  slightly  flexed  in  the  dissepimen- 
tarium.  The  major  septa  are  dilated  up  to  0-35  mm.  across  for  a  length  of  o*6  to  1 
mm.  just  outside  the  tabularium.  The  minor  septa  are  only  slightly  thickened  in 
the  same  area.  The  latter  do  not  penetrate  the  tabularium  but  the  major  septa, 
extremely  attenuated,  extend  usually  to  within  0-4  mm.  of  the  axis  where  they  are 
usually  curved  into  a  very  faint  vortex.  There  is  no  dilatation  of  the  axial  ends  of  the 
major  septa.  Carinae  are  not  developed  and  the  septa  are  usually  smooth  through- 
out. 

Dissepiments  are  usually  uniserial  between  adjacent  septa.  The  trace  of  the 
horseshoe  dissepiments,  corresponding  in  position  with  the  septal  dilatation,  forms  a 
well  marked  wall  around  the  tabularium. 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  series  of 
variably  shaped  dissepiments,  from  near  flat  to  quite  globose.  Just  outside  the 
tabularium,  the  dissepimental  surface  rises  to  a  crest  formed  by  a  somewhat  dis- 
continuous vertical  series  of  horseshoe  dissepiments.  Steep  sloping  vesicles  are 
occasionally  developed  between  the  horseshoe  dissepiments  and  the  tabularium. 

The  structure  of  the  tabularium  is  simple.  Tabulae  are  incomplete,  distally 
concave  plates  or  low  elongate  vesicles.  They  are  arranged  to  give  a  flat  or  slightly 
concave  surface  to  the  tabularium. 

The  few  data  available  for  the  species  are  summarized  in  Table  6. 

Discussion.  P.  rozkowskae  is  probably  the  species  of  Phillipsastrea  which  most 
closely    approaches    Frechastraea.     Its    septa    are    very    similar    to    the    typical 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  231 

frechastraeid  form  (see  p.  232)  and  the  series  of  horseshoe  dissepiments  is  relatively 
poorly  developed  for  Phillipsastrea. 

P.  rozkowskae  is  similar  in  its  pseudocerioid  growth  form  to  P.  smithi  (Rozkowska), 
P.  ananas  (Goldfuss)  and  P.  macrommata  (F.  A.  Roemer).  It  differs  from  the  latter 
two,  however,  through  its  smaller  tabularium  size  and  lower  At/A  ratio  as  well  as 
the  character  of  its  septal  thickening.  P.  smithi  is  distinguished  by  very  strong 
septal  dilatation  forming  a  wide,  dense  wall  around  the  tabularium.  On  the  axial 
side  of  this  wall,  the  major  septa  are  represented  by  thick  lobes  only,  and  do  not 
cross  the  tabularium  towards  the  axis. 


Genus  FRECHASTRAEA  nov. 

Derivation  of  name.     The  genus  is  named  after  Fritz  Freeh  (1861-1917). 

Diagnosis.  Massive  rugose  corals,  pseudocerioid,  astraeoid  or  thamnasterioid. 
Corallites  small  with  clearly  defined  wall  at  tabularium  junction.  Septa  of  two 
orders,  major  and  minor,  generally  slightly  thickened  against  tabularium  junction 
and  strongly  attenuate  in  tabularium.  In  longitudinal-section,  dissepiments  small, 
globose.  Septal  trabeculae  arranged  in  a  fan  usually  on  the  series  of  dissepiments 
adjacent  to  tabularium.  Horseshoe  dissepiments  may  rarely  occur  in  this  series. 
Dissepimental  surface  almost  flat  with  a  slight  elevation  surrounding  tabularium. 
Tabulae  complete  or  incomplete. 

Type  species.  Cyathophyllum  pentagonum  Goldfuss  (1826  :  60,  pi.  19,  fig.  3). 
Frasnian;  Namur,  Belgium. 

Distribution.     Widespread  in  the  Frasnian  of  Europe. 

Other  species  assigned  to  Frechastraea.  Acervularia  goldfussi  de  Verneuil  & 
Haime  (1850  :  161);  ISmithia  boloniensis  Edwards  &  Haime  (1851  :  423);  Smithia 
bowerbanki  Edwards  &  Haime  (1851  :  423);  Smithia  micrommata  C.  F.  Roemer 
(1852  :  197,  pi.  51,  figs.  20a,  b);  ? Acervularia  roemeri  var.  J3  concinna  F.  A.  Roemer 
(1855  :  32,  pi.  6,  figs.  iqa~c);  Phillipsastraea  sanctacrucensis  Rozkowska  (1953  :  59, 
text-figs.  32-33,  pi.  2,  figs.  8-10) ;  Phillipsastraea  pentagona  (Goldfuss)  var.  minima 
Rozkowska  (1953  :  66,  text-figs.  36-38,  pi.  8,  fig.  9). 

Discussion.  Frechastraea  embraces  a  well  defined  group  of  Devonian  colonial 
rugose  corals.  Species  here  included  in  the  genus  have  been  previously  assigned  to 
Hexagonaria,  Billingsastraea,  Phillipsastrea  or  synonyms  of  these  three  genera. 

Workers  in  the  last  century,  particularly  Edwards  &  Haime,  frequently  placed 
pseudocerioid  and  astraeoid  Devonian  corals,  including  species  of  Frechastraea,  in 
the  Silurian  genus  Acervularia  Schweigger.  This  similarity  between  the  Silurian 
and  Devonian  forms  is,  however,  quite  superficial  and  Acervularia  has  long  been 
considered  unrelated  to  the  phillipsastreids.  More  or  less  concurrent  with  the  use 
of  Acervularia,  some  astraeoid  and  thamnasterioid  species  of  Frechastraea  were 
assigned  to  Smithia,  which  is  an  objective  synonym  of  Phillipsastrea. 

Lang  &  Smith  (1935  :  559)  referred  to  Prismatophyllum  pentagona  (Goldfuss)  in 
their  remarks  on  the  genus  Prismatophyllum.    Prismatophyllum  Simpson  is  con- 


232  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

sidered  a  junior  synonym  of  Hexagonaria  Gurich  (see  Lang,  Smith  &  Thomas  1940  : 
104).  The  type  species  of  both,  P.  prisma  Lang  &  Smith  and  H.  hexagona  (Goldfuss) 
respectively,  are  characterized  by  typical  disphyllid  dissepimentaria  with  the  septal 
trabeculae  sloping  axially  and  upwards  throughout.  Frechastraea  differs  from  these 
chiefly  by  the  possession  of  a  full  trabecular  fan  based  on  the  series  of  dissepiments 
adjacent  to  the  tabularium.  In  addition,  Hexagonaria  shows  no  tendency  to  the 
development  of  horseshoe  dissepiments  and  lacks  the  strongly  defined  tabularium 
junction  characteristic  of  the  present  genus. 

Many  of  the  forms  assigned  to  Hexagonaria  by  Moenke  (1954)  also  do  not  possess 
a  disphyllid  dissepimentarium.  These  species,  however,  are  morphologically  quite 
distinct  from  those  referred  to  Frechastraea  and  have  been  placed  in  Marisastrum 
by  Rozkowska  (1965  :  262)  and  Scrutton  (1967  :  270). 

Schouppe  (1958  :  235)  included  species  of  Frechastraea  in  the  genus  Billingsastraea 
Grabau.  The  type  species  of  Billingsastraea  is  B.  verneuili  (Edwards  &  Haime). 
The  holotype  is  missing  but  Ehlers  &  Stumm  (1953  :  2)  have  described  corals  under 
that  name  which  appear  in  all  respects  to  satisfy  the  original  diagnosis.  Billingsas- 
traea sensu  stricto  is  thus  characterized  by  a  broadly  reflexed  dissepimentarium  in 
which  there  is  no  tendency  to  produce  horseshoe  dissepiments.  The  septa  are 
uniformly  thin  from  the  periphery  to  the  axis  of  the  corallite,  usually  with  no  sign 
of  dilatation,  and  the  strongly  marked  tabularium  junction  of  Frechastraea  is  lacking. 
Oliver  (1964  :  2-3)  criticized  Schouppe's  interpretation  of  Billingsastraea  and 
suggested  that  there  is  probably  no  close  relationship  between  that  genus  and  the 
philhpsastraeids.  The  writer  subscribes  to  this  view  (Scrutton  1967  :  276)  and 
species  of  Frechastraea  are  here  considered  quite  distinct  from  Billingsastraea. 

Schouppe  (1958  :  156)  recorded  a  well  developed  series  of  horseshoe  dissepiments  in 
P.  hennahi,  the  type  species  of  Phillipsastrea.  Until  that  time,  species  of  Frechastraea 
had  been  widely  assigned  to  that  genus.  They  are  certainly  most  closely  related  to 
Phillipsastrea  but  may  be  distinguished  from  it  by  the  rarity  with  which  horseshoe 
dissepiments  occur.  Frechastraea  is  also  strongly  characterized  by  its  septa  (compare 
Text-figs.  46  &  4c) — uniformly  thick  in  the  dissepimentarium  with  a  short  club- 
shaped  thickening  against  the  tabularium  boundary,  and  major  septa  strongly 
attenuate  in  the  tabularium.  In  Phillipsastrea  the  septa  are  more  spindle-shaped, 
and  may  be  equally  thin  in  the  tabularium  and  peripheral  parts  of  the  dissepimen- 
tarium; frechastraeid  type  septa  are  only  rarely  known.  Finally  species  of  Frechas- 
traea tend  to  have  smaller  individuals — seldom  greater  than  2mm.  tabularium 
diameter — than  species  of  Phillipsastrea. 

The  occasional  horseshoe  dissepiments  developed  in  the  dissepimentaria  of  species 
of  Frechastraea  are  recorded  here  for  the  first  time.  It  is  thought  that  the  genus 
descended  from  a  phillipsastreid  ancestor  chiefly  through  the  progressive  elimination 
of  horseshoes  from  the  dissepimentarium.  P.  rozkowskae,  in  particular,  shows  a 
tendency  toward  frechastraeid  characters  and  is  considered  a  possible  intermediary 
between  the  two  genera.  It  may  be  significant  in  an  evolutionary  sense  that 
Rozkowska  (1953)  did  not  record  horseshoes  in  species  of  Frechastraea  from  mainly 
Upper  Frasnian  horizons  in  Poland  whilst  they  can  often  be  seen  in  the  dissepi- 
mentaria of  the  same  species  among  the  English  Lower  Frasnian  material. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  233 

Frechastraea  pentagona  pentagona  (Goldfuss) 

Plate  6,  fig.  5;   Plate  7,  figs.  1-5 

1826  Cyathophyllum  penlagonum  Goldfuss  :  60,  pi.  19,  fig.  3. 

1840  Astrea  (Favastrea)  pentagona  (Goldfuss)  Lonsdale  :  697  pars,  pi.  58,  fig.  1  (non  la). 

1841  Astrea  pentagona  (Goldfuss);    Phillips  :  11,  pi.  6,  fig.  15. 
1851  Acervularia  pentagona  (Goldfuss)  Edwards  &  Haime  :  418. 

1853  Acervularia  pentagona  (Goldfuss);    Edwards  &  Haime  :  238,  pi.  53,  figs.  5-56. 

1883  Acervularia  pentagona  (Goldfuss);   C.  F.  Roemer  :  352,  text-fig.  70. 

1885  Phillipsastrea  pentagona  (Goldfuss)  Freeh  :  54  pars,  pi.  3,  figs.  7,  ja,  ?8,  10;   pi.  8,  fig.  3. 

1935  Prismatophyllum  pentagona  (Goldfuss)  Lang  &  Smith  :  559. 

1953  Phillipsastraea  pentagona  (Goldfuss);    Rozkowska  :  64,  text-figs.  36—39,  pi.  8,  fig.  7. 

But  not: 

1952     Phillipsastraea  pentagona  (Goldfuss) ;    Soshkina  :  102,  pi.  43,  fig.  145. 

Diagnosis.  Pseudocerioid  Frechastraea  with  mean  tabularium  diameter  1-09  mm. 
and  7  to  13  major  septa  (East  Ogwell  sample).  Smooth,  non-carinate  septa. 
Horseshoe  dissepiments  very  rarely  developed.  Tabulae  complete  or  incomplete. 
Increase  axial  or  lateral. 

Lectotype  (selected  by  J.  W.  Pickett  in  press).  Original  of  Goldfuss  1826,  pi. 
19,  fig.  3  which  is  specimen  206  in  the  Goldfuss  Collection  of  the  Geologisch-Palaon- 
tologisches  Institut,  Bonn.  Goldfuss  (1826  :  60)  gave  the  horizon  and  locality  as 
the  "  Transition  limestone  of  the  Namur  region  ",  Belgium.     Frasnian. 

Material.  Ramsleigh  Quarry:  OUM  D279  (Colony  1),  BM(NH)  R23400 
(Colony  3),  BM(NH)  R5635  (Colony  4).  Other  measured  specimens:  BM(NH) 
R676,  TM(JB)  305A. 

Road  cutting,  20  yd.  west  of  Ramsleigh  Quarry  entrance:  OUM  D538  (Colony  2). 
Other  measured  specimen:   OUM  D537. 

Road  cutting,  40  yd.  west  of  Ramsleigh  Quarry  entrance.  Measured  specimen : 
OUM  D532. 

Distribution.  England:  Lower  Frasnian  limestones,  Ramsleigh  Quarry  and 
adjacent  road  cutting,  East  Ogwell,  near  Newton  Abbot,  south  Devon.  Also 
Frasnian  of  Belgium,  Germany  and  Poland. 

Description.  Colony  shape,  external  features  and  overall  dimensions  are  unknown 
as  all  the  English  material  is  fragmentary.  Rozkowska  (1953  :  64),  however, 
described  Polish  representatives  of  this  subspecies  as  thick,  tabular  colonies  up  to 
6  cm.  in  diameter  and  6  cm.  in  height.  Goldfuss'  original  specimen  is  a  thick, 
rectangular  block  with  the  convex  upper  surface  measuring  7  cm.  by  8  cm.  The 
corallites  have  slightly  depressed  tabularia  with  a  low,  encircling  ridge  formed  by  the 
innermost  series  of  dissepiments.  Rozkowska  gave  no  details  of  a  holotheca  and 
it  is  not  preserved  in  any  of  the  English  specimens. 

The  colonies  are  massive,  pseudocerioid,  tending  rarely  to  astraeoid  or  thamnas- 
terioid.  Individual  corallites,  usually  pentagonal  or  hexagonal  in  shape,  are  separated 
from  each  other  by  a  straight  or  zigzagged  pseudotheca.  Occasionally  the  pseudo- 
theca  may  break  down  when  the  septa  are  more  or  less  confluent  from  one  corallite 
to  the  next. 


234 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


40    r 


f      20      ■ 


f      20 


Colony  2 


Total   sample 


12    r 


11      • 


n     10    • 


08 


10  12 

at        (mm) 


12 
11 
10 


%      > 


16 


08  10  12 

at        (mm) 


Total  sample  Colony        1  Colony      2  Colony      3 

Fig.  13.     Frechastraea  pentagona  pentagona. 


14 


Colony       4 


16 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


235 


Table  7. — Statistical  data  for  some  characters  of  Frechastraea  penlagona  pentagona. 

Total  sample        Colony  1  Colony  2  Colony  3  Colony  4 

N  424  (8)  60  60  60  60 


O.R. 

0-8-1-5 

0 • 9-1 • 1 

0-9-1-3 

I -O-I -2 

1-0-1-3 

X 

1  -09 

1  -oo 

1  -04 

I  -09 

I  -20 

dt 

s 

o- 10 

0-059 

0-084 

0-067 

0-068 

C.V. 

9-38 

5-97 

8-13 

6-12 

5-65 

S.E.m 

0-005 

0-0077 

O-OII 

0-0086 

0-0087 

O.R. 

7-12 

7-i  1 

8-1 1 

9-12 

0-12 

X 

IO-I2 

9-80 

9-73 

IO-I2 

9-90 

n 

s 

0-32 

0-38 

030 

0-15 

0-24 

C.V. 

3-15 

3-85 

3-°4 

1-48 

2-43 

S.E.m 

0-016 

0-049 

0-038 

0-019 

0-031 

O.R. 

6-67-12 -oo 

7-78-11 -ii 

7-69-10-00 

8 • 33-10 • 00 

7-50-10-00 

X 

9-36 

9-86 

9-73 

9-32 

8-30 

n/dt 

s 

o-59 

0-25 

0-30 

0-44 

o-33 

C.V. 

6-34 

2-54 

3-04 

4-69 

3-93 

S.E.m 

0-029 

0-032 

0-038 

0-056 

0-042 

Graphs  :— 

r 

0-92 

094 

0-74 

099 

0-91 

n/dt 

a 

3-13 

6-35 

3-51 

2-25 

3-57 

b 

6-72 

3H9 

6-09 

7-67 

5-63 

n/dt/ 
/td 

r 

a 

—  1  -oo 

-5-8i 

—  0-90 

-4-21 

—  0-96 

—  6-io 

-o-99 
-6-56 

-0-94 

-4-83 

/ 

b 

15-68 

i4-°5 

15-75 

16-42 

14-07 

A 

At 

At/A 

Graph:— At/A 

O.R. 

0-091- 

-0-16         o-oc 

178-0-011         o- 

057-0-099 

X 

O' 

12 

0 

■0093 

0-079 

r 

0-38 

s 

O' 

024 

0 

•0013 

0-015 

a 

0-055 

C.V. 

20  ■ 

25 

14 

•38 

18-63 

b 

0-0027 

S.E.m 

O' 

0086 

0 

■00047 

0-0052 

The  septa,  both  major  and  minor,  are  straight  and  uniformly  thick,  about  0-05  mm. 
across,  in  the  dissepimentarium.  At  the  tabularium  junction  the  major  septa,  and 
to  a  lesser  degree  the  minor,  are  dilated  to  form  a  short  bulbous  thickening.  This 
thickening,  associated  with  the  clearly  defined  junction  between  tabulae  and  dissepi- 
ments, gives  rise  to  a  strong  internal  wall  at  this  point.  The  minor  septa  are  normally 
confined  to  the  dissepimentarium  and  only  occasionally  show  as  slight  projections 
into  the  tabularium.  The  major  septa,  however,  continue  into  the  tabularium  as 
extremely  thin  processes,  usually  extending  halfway  or  slightly  more  towards  the 
axis  and  sometimes  reaching  the  axis  itself.  There  is  a  strong  tendency  for  the  axial 
ends  of  the  longer  major  septa  to  fuse.     The  thin  septal  elements  in  the  tabularium 


236  COLONIAL  PHI  LLIPS  ASTR  AEI  D  A  E  FROM  S.E.  DEVON 

are  easily  obliterated  by  slight  recrystallization,  but  even  in  the  best  of  the  material 
examined  there  was  no  sign  of  thickening  of  the  axial  ends  of  the  major  septa. 

The  dissepiments  are  almost  always  uniserial  between  adjacent  septa  and  appear 
moderately  spaced  in  cross-section.  In  longitudinal-section,  the  dissepimentarium 
is  composed  of  several  series  of  small,  evenly  developed,  well  arched  dissepiments. 
Their  vertical  height  varies  between  o-i  and  o-2  mm.  The  dissepimental  surface  is 
usually  flat  but  may  occasionally  slope  slightly  away  from  the  tabularium  boundary. 
At  the  latter  position,  a  dissepiment  may  very  rarely  be  modified  into  a  horseshoe 
shape  but  there  is  no  tendency  to  develop  a  series  of  horseshoe  dissepiments.  No 
more  than  two,  or  possibly  three  horseshoes  have  been  observed  in  continuous  vertical 
succession. 

The  tabularium  structure  is  often  partially  obscured  by  slight  recrystallization. 
The  tabulae  are  usually  flat  lying  and  may  be  complete  or  interleaved.  Occasionally, 
however,  they  slope  downwards  away  from  from  the  axis  with  an  overall  tent  or  bell 
shaped  appearance.  In  some  cases  there  is  evidence  of  axial  structures  suggesting 
the  vesicles  described  in  the  tabularium  of  Frechastraea  goldfussi  (see  p.  249). 
Unfortunately,  poor  preservation  and  sometimes  septal  traces  obscure  the  axial  area 
at  these  points  and  the  relationships  of  the  tabulae  are  not  clear.  In  the  more 
simply  constructed  tabularia,  the  tabulae  have  an  average  vertical  spacing  of  about 
0*2  mm. 

Both  axial  and  lateral  increase  have  been  observed  in  this  subspecies,  the  latter 
being  more  common. 

A  sample  of  eight  colonies  from  the  limestones  exposed  in  Ramsleigh  Quarry  and 
the  adjacent  road  cutting  has  been  statistically  analysed.  Four  of  the  colonies  have 
also  been  analysed  individually.  The  statistics  are  listed  in  Table  7  and  illustrated 
graphically  in  Text-figs.  13  and  16. 

Discussion.  Both  Lonsdale  (1840  :  697)  and  Freeh  (1885  :  54)  included  Acer- 
vularia  goldfussi  or  specimens  belonging  to  this  species  in  their  interpretation  of 
Cyathophyllum  pentagonum.  The  two  species  are  superficially  extremely  similar, 
although  on  close  inspection  they  can  be  readily  distinguished.  The  details  of  the 
differences  are  given  below  (p.  253)  under  F.  goldfussi. 

The  most  recent  thorough  investigation  of  Frechastraea  pentagona  pentagona  is  by 
Rozkowska  (1953  :  64)  as  a  species  of  Phillipsastrea.  Her  material  has  a  slightly 
larger  mean  tabularium  diameter  than  the  English  sample  but  there  is  no  doubt  that 
they  belong  to  the  same  subspecies. 

Frechastraea  pentagona  (Goldfuss)  minima  (Rozkowska) 
Plate  8,  figs,  r-3 

1953     Phillipsastraea  pentagona  (Goldfuss)  var.  minima  Rozkowska  :  66,  text-figs.  36-38,  pi.  8, 

fig-  9- 
1953     Phillipsastraea  bowevbanki  (Edwards  &  Haime) ;   Rozkowska  :  67,  pi.  8,  figs.  3,  4. 
*959     Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer) ;    Middleton  :  156, 

text-fig.  6/. 

Diagnosis.  Pseudocerioid  to  astraeoid  to  thamnasterioid  Frechastraea.  Mean 
tabularium  diameter  0-96  mm.  with  7  to  12  major  septa  (East  Ogwell  sample). 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  237 

Tabularia  somewhat  unevenly  spaced.     Septa  sinuous  and  only  slightly  dilated  at 
tabularium  junction.     Horseshoe  dissepiments  rare.     Increase  axial  or  lateral. 

Holotype.  A98  in  the  Collections  of  the  Polska  Akademia  Nauk,  Pracownia 
Palaeozoologii,  Poznan,  Poland.     Upper  Frasnian,  Psie  Gorki,  Kielce,  Poland. 

Material.  Ramsleigh  Quarry:  GSM  73119  (Colony  1),  GSM  11822  (Colony  2), 
GSM  73118  (Colony  3).  Other  measured  specimens:  BM(NH)  R46371,  BM(NH) 
R23266.     Additional  material :   GSM  11821,  GSM  11823. 

Distribution.  England:  Lower  Frasnian  of  Ramsleigh  Quarry,  East  Ogwell, 
near  Newton  Abbot,  south  Devon.     Also  Upper  Frasnian  of  Kielce,  Poland. 

Description.  Details  of  size,  shape  and  external  features  of  the  English  material 
are  unknown  as  it  is  fragmentary.  Rozkowska  (1953  :  67),  however,  described  the 
holotype  as  a  low,  plate-like  colony  measuring  3*5  cm.  by  i*6  cm.  in  surface  area. 

Individual  corallites  are  seldom  completely  surrounded  by  an  external  pseudotheca 
and  usually  the  septa  are  confluent  or  subconfluent  between  them.  All  gradations 
from  the  pseudocerioid  to  the  thamnasterioid  stage  are  usually  present  in  the  same 
colony  and  this  is  accompanied  by  a  general  increase  in  the  distances  separating  the 
tabularia.  Where  an  external  pseudotheca  is  present,  it  is  formed  by  the  deflection 
of  the  peripheral  septal  ends. 

The  septa,  0-05  mm.  or  less  in  thickness,  are  uniformly  thin  and  are  gently  sinuous 
between  tabularia.  They  are  usually  smooth  sided.  Rarely  there  appears  to  be  a 
slight  separation  of  the  trabeculae,  which  remain  unthickened,  resulting  in  dis- 
continuous septa.  At  the  tabularium  junction,  the  septa  are  slightly  dilated  for 
about  ot  mm.  of  their  length.  Within  the  tabularium,  the  major  septa  thin  abruptly 
and  continue  as  very  fine  filaments  more  or  less  to  the  axis,  where  the  ends  of  two 
or  more  adjacent  septa  may  fuse.  The  minor  septa  do  not  penetrate  into  the 
tabularium. 

The  dissepiments  are  almost  always  uniserial  between  adjacent  septa.  The 
tabularium  junction  is  sharply  defined  in  cross-section,  giving  the  appearance  of  an 
internal  wall. 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  series  of 
small,  well  arched  dissepiments.  They  may  vary  somewhat  in  size  and  their  height, 
usually  about  o-i  mm.,  ranges  from  0*05  to  0*25  mm.  The  surface  of  the  dissepi- 
mentarium is  flat  peripherally,  usually  rising  slightly  with  the  more  globose  dissepi- 
ments adjacent  to  the  tabularium.  Horseshoe  dissepiments  may  rarely  develop  in 
the  latter  position. 

The  tabularium  structure  is  simple  and  the  tabulae  maybe  complete  or  incomplete. 
The  complete  tabulae  are  either  flat  or  sag  slightly  in  the  middle.  The  incomplete 
tabulae  are  flat  to  slightly  bowed  and  interleaved  with  each  other,  or  in  the  form  of 
long  weakly  arched  vesicles.  The  vertical  spacing  of  the  tabulae  may  vary  between 
o-i  and  03  mm.,  but  is  normally  about  0*15  mm.  in  the  more  regularly  developed 
tabularia. 

Examples  of  both  axial  and  lateral  increase  have  been  observed. 

A  statistical  analysis  has  been  made  of  a  total  sample  of  5  colonies  from  Ramsleigh 


238  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

100    r 


90     • 


JO      ■ 


70      • 


60 


f        50      • 


40      • 


30     ■ 


20     ■ 


10 


06  07 


10 


0  9 

10 

dt 

(mm) 

13 

12 

11 

n/     in 

A  10 

9 

8 

7 

06 


08  10  1-2 

at         (mm) 


14  06 


08  10 

dt         (mm) 


14 


12  14 


Total    sample 


Colony     I  Colony     2  Colony     3 

Fig.  14.     Frechastraea  pentagona  minima. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  239 

Table  8. — Statistical  data  for  some  characters  of  Frechastvaea  pentagona  minima. 


Total  sample 

Colony  1 

Colony  2 

Colony  3 

N 

24°  (5) 

60 

60 

60 

O.R. 

0-7-1-3 

0-7-1 -i 

0 ■ 8-1 • 1 

0-9-1  -r 

X 

0-96 

0-90 

o-93 

0-98 

dt 

s 

O-II 

0-079 

0-082 

0-055 

C.V. 

10-98 

8-78 

8-85 

5-57 

S.E.m 

0-0068 

O-OIO 

OOII 

0-0071 

O.R. 

7-12 

8-IO 

7-1 1 

8-10 

X 

9-54 

9-61 

9-25 

9-52 

n 

s 

0-42 

0-23 

o-54 

0-36 

C.V. 

4-42 

2-42 

5-86 

3-77 

S.E.m 

0-027 

0-030 

0-070 

0-046 

O.R. 

7-69-12 -87 

8-18-12-87 

7-78-11 -26 

8-89-II -12 

X 

10-05 

10-72 

IO-OI 

9.72 

n/dt 

s 

0-64 

o-73 

o-35 

0-20 

C.V. 

6-40 

6-82 

3-54 

2-04 

S.E.m 

0-42 

0-094 

0-046 

O-026 

Graphs: — 

r 

0-98 

0-77 

o-95 

0-97 

n/dt 

a 

4-00 

2-94 

6-6i 

6-58 

b 

5-71 

6-96 

3-12 

3-07 

n/dt/ 
/dt 

r 
a 

—  0-98 
-6-ii 

-0-99 
-9-24 

-o-88 
-4-32 

—  0-92 
-3-63 

/ 

b 

I5-9I 

19-06 

14-01 

13-27 

A 

At 

At/A 

Graph : 

—At/A 

O.R. 

0 • 1 1-0 ■ 1 

5     0-0065-0-010     0-054-0-071 

X 

0-12 

00075 

0-062 

r 

0-84 

s 

0-OI7 

0-0016 

0-0076 

a 

0-092 

C.V. 

I4-52 

21-37 

12-21 

b      - 

-0-0036 

S.E.m 

O-O07! 

3           0-00072 

OOO34 

Quarry.     Three  of  these  colonies  have  also  been  analysed  individually.     The  statistics 
are  listed  in  Table  8  and  illustrated  graphically  in  Text-figs.  14  and  16. 

Discussion.  Rozkowska  (1953  :  66)  erected  this  subspecies  as  a  variety  of 
"  Phillipsastraea  pentagona  "  on  only  one  specimen.  Although  the  tabularia  of  her 
specimen  (0 -6-0-8  mm.  diameter)  are  slightly  smaller  than  those  in  the  English 
examples,  the  colonies  are  in  complete  morphological  agreement  and  are  undoubtedly 
the  same  subspecies.  The  holotype  clearly  shows  the  characteristic  pseudocerioid 
grading  to  thamnasterioid  nature  of  the  coralhtes  and  the  irregular  spacing  of  the 
tabularia  described  above. 


240  COLONIAL  PHILLI  PS  ASTR  AEI  D  AE  FROM  S.E.  DEVON 

The  specimen  described  by  Rozkowska  (1953  :  67)  as  "  Phillipsastraea  bowerbanki  " 
is  indistinguishable  from  GSM  73118  (PL  8,  fig.  1)  of  the  present  material  and  is 
thus  also  referable  to  this  subspecies.  The  slight  morphological  differences  between 
these  specimens  and  the  holotype  of  F.  pentagona  minima  lie  within  the  range  of 
variation  to  be  expected.  F.  pentagona  minima  is  quite  distinct  from  F.  bowerbanki 
which  is  almost  exclusively  thamnasterioid  and  has  considerably  larger  tabularia, 
wide  and  regularly  spaced  (see  p.  253). 

Two  fragments  from  Ramsleigh  Quarry  (GVM  26/7  and  26/8),  misidentified  as 
Phillipsastraea  pentagona  var.  micrommata  by  Middleton  (1959  :  156),  also  belong  to 
this  subspecies. 

F.  pentagona  pentagona  and  F.  pentagona  minima  are  very  similar  in  quantitative 
terms.  Only  0*13  mm.  separates  their  mean  tabularium  diameters.  In  addition, 
their  lines  on  the  graphs  plotting  septal  number  against  tabularium  diameter  and 
tabularium  area  against  corallite  area  are  very  similar  in  both  slope  and  position 
(Text-fig.  16).  Nevertheless,  because  of  the  large  sample  sizes,  tests  of  statistical 
discrimination  show  the  subspecies  to  differ  significantly  in  both  dt  (Table  ro)  and 
their  lines  on  the  former  of  the  two  graphs. 

F.  pentagona  minima  is  distinguished  principally  on  qualitative  characteristics,  and 
without  prior  separation  on  these  grounds,  could  not  be  differentiated  from  F. 
pentagona  pentagona  on  quantitative  data  alone.  F.  pentagona  pentagona  is  pseudo- 
cerioid,  tending  slightly  to  astraeoid,  with  straight  septa.  The  septa  in  F.  pentagona 
minima,  on  the  other  hand,  are  generally  sinuous  and  associated  with  a  wide  range  in 
form  from  pseudocerioid  to  thamnasterioid.  The  tabularia  also  tend  to  be  irregularly 
spaced.  In  the  material  so  far  examined,  no  difficulty  has  been  found  in  distinguish- 
ing quite  clearly  between  the  two  subspecies. 

Both  occur  in  association  in  the  Lower  Frasnian  Ramsleigh  limestones  in  England 
and  the  Upper  Frasnian  of  Kielce  in  Poland  (Rozkowska  1953).  In  the  English 
fauna,  F.  pentagona  pentagona  and  F.  pentagona  minima  occur  in  approximately 
equal  numbers  whilst  Rozkowska's  figures  suggest  a  ratio  of  11  :  1  in  favour  of  F. 
pentagona  pentagona  in  the  Kielce  area. 

Frechastraea  micrommata  (C.  F.  Roemer) 

Plate  8,  figs.  4,  5 

1852     Smithia  micromata  C.  F.  Roemer  :  197,  pi.  51,  figs.  20a,  b. 

But  not: 

1885     Phillipsastrea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer)  Freeh  :  56,  pi.  3,  figs 

11-13;    pi.  8,  fig.  1. 
1953     Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer) ;    Rozkowska  :  66, 

text-figs.  36-38;   pi.  8,  fig.  8. 
1958     Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer);    Bulvanker  :  123, 

pi.  60,  figs.  1-3;    pi.  61,  figs.  4,  5. 
!959     Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer);    Middleton  :  156, 

text-fig.  6/. 

Diagnosis.  Pseudocerioid  tending  to  astraeoid  Frechastraea.  Mean  tabularium 
diameter  1*42  mm.  with  18  to  21  septa  (holotype).     Major  and  minor  septa  not 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  241 

distinguished  as  septa  penetrating  the  tabularium  exceedingly  rare;  septa  non- 
carinate.  Tabularium  structure  very  simple,  usually  composed  of  complete  tabulae. 
Increase  lateral. 

Lectotype  (here  chosen).  Specimen  no.  34  in  the  collections  of  the  Geologisch- 
Palaontologisches  Institut,  Bonn.  Original  of  Roemer's  (1852,  pi.  51)  figures  20a,  b. 
Frasnian ;  Ferques  near  Boulogne,  France. 

Description.  Colony  irregularly  disc  shaped,  8  cm.  in  diameter  and  4  cm.  high; 
covered  basally  by  a  strongly  concentrically  ridged  holotheca.  The  corallites  are 
pseudocerioid  tending  to  astraeoid  with  a  thin,  usually  zigzagged,  pseudotheca. 

The  septa  are  not  normally  distinguished  as  major  and  minor  as  with  very  rare 
exceptions  they  all  end  at  the  tabularium  junction.  The  septa  are  about  0-05  mm. 
thick  in  the  dissepimentarium  and  follow  a  straight  or  slightly  sinuous  course  to  the 
tabularium  boundary  where  they  may  thicken  slightly.  Septa  may  occasionally 
form  slight  inward  projections  on  the  tabularium  boundary  and  one  septum  was 
seen  to  penetrate  the  tabularium  for  a  distance  of  0*4  mm. 

The  tabularium-dissepimentarium  junction  is  a  thin  strong  circular  wall  in  cross- 
section.  This  wall  appears  to  be  the  product  of  septal  dilatation  and  thickening 
of  the  inner  arms  of  the  series  of  dissepiments  adjacent  to  the  tabularium.  Some- 
times, however,  the  septa  themselves  appear  to  bend  sharply  at  the  boundary  to 
form  a  segment  of  the  wall. 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  series  of  quite 
globose  dissepiments  normally  0 '5-0*6  mm.  in  height.  In  narrow  levels  which  can 
be  traced  from  corallite  to  corallite,  however,  they  become  smaller,  flatter  and  more 
closely  spaced.  The  dissepimental  surface  is  flat  over  most  of  its  area  but  rises 
slightly  to  a  crest  adjacent  to  the  tabularium.  Horseshoe  dissepiments  have  not 
been  observed. 

Tabularium  structure  is  very  simple,  consisting  in  the  main  of  complete  flat  or 
slightly  bowed  tabulae  with  occasional  incomplete  arched  plates.  The  latter  may 
be  subsidiary  to  the  complete  tabulae  or  may  interleaf  to  form  the  tabularium 
structure  proper. 

One  example  of  lateral  increase  has  been  observed. 

Measurements  on  twelve  corallites  showed  tabularium  diameter  to  range  from 
I'3-I*5  mm.,  mean  value  1-42  mm.,  with  18  to  21  septa.  Mean  n/dt,  taking  n  to  be 
half  the  number  of  septa  in  each  corallite,  is  6*94. 

Discussion.  The  above  description  is  based  on  Roemer's  figured  specimen  only. 
Although  no  horseshoe  dissepiments  have  been  observed,  the  character  of  the  species 
as  a  whole  places  it  without  doubt  in  Frechastraea.  Full  knowledge  of  the  variation 
in  dissepimental  shape  must  await  the  description  of  further  material  belonging  to 
this  species. 

This  is  the  first  time  that  the  type  specimen  has  been  sectioned  and  the  slides  show 
that  Freeh  (1885  :  56)  and  subsequent  authors  have  misidentified  the  species.  The 
specimens  placed  by  Freeh  and  others  in  Smithia  micrommata  (as  Phillipsastrea 
pentagona  var.  micrommata)  are  here  assigned  toF.  carinata  sp.  nov.  which  is  described 
below. 


242  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

F.  micrommata  differs  from  all  other  known  species  of  Frechastraea  by  the  lack 
of  any  distinction  between  major  and  minor  septa.  In  particular,  it  can  be  dis- 
tinguished from  F.  carinata  by  the  latter's  septal  carination  and  more  complex 
tabularium  structure.  Data  so  far  available  show  that  tabularia  in  F.  micrommata 
are  somewhat  larger  than  in  F.  carinata  (dtx  =  i>02  mm.). 

Frechastrea  carinata  sp.  nov. 
Plate  9,  figs.  1-3 

1885     Phillipsastrea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer)  Freeh  :  56,  pi.  3, 

figs.  11-12,  ?i3;    pi.  8,  ?fig.  1. 
1953     Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer) ;    Rozkowska  :  66, 

text-figs.  36-38,  pi.  8,  fig.  8. 
But  not: 
1852     Smithia  micrommata  C.  F.  Roemer  :  197,  pi.  51,  figs.  20a,  b. 

1958  Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer);  Bulvanker  :  123, 
pi.  60,  figs.  1-3;   pi.  61,  figs.  4-5. 

1959  Phillipsastraea  pentagona  (Goldfuss)  var.  micrommata  (C.  F.  Roemer);    Middleton  :  156, 
text-fig.  6/. 

Derivation  of  name.  The  name  refers  to  the  presence  of  septal  carinae,  an 
important  distinguishing  feature  of  this  species. 

Diagnosis.  Pseudocerioid,  tending  to  astraeoid  and  thamnasterioid  Frechastraea. 
Mean  tabularium  diameter  1-02  mm.  and  between  8  and  14  major  septa  (topotype 
sample).  Septa  variably  but  typically  carinate.  Horseshoe  dissepiments  very 
rare.     Tabulae  mainly  incomplete.     Increase  axial  or  lateral. 

Holotype.  OUM  D309.  Lower  Frasnian ;  road  cutting,  80  yd.  west  of  Rams- 
leigh  Quarry  entrance,  East  Ogwell,  near  Newton  Abbot,  south  Devon. 

Material.  Ramsleigh  Quarry:  BM(NH)  R23210  (Colony  1),  BM(NH)  R23211 
(Colony  2),  BM(NH)  R23216  (Colony  3),  BM(NH)  R5640  (Colony  4).  Other  measured 
specimens:   BM(NH)  R677,  BM(NH)  R5634. 

Road  cutting,  80  yd.  west  of  Ramsleigh  Quarry  entrance.  Measured  material : 
OUM  D309-310,  OUM  D535-6. 

Distribution.  England:  Lower  Frasnian  limestones,  Ramsleigh  Quarry  and 
the  adjacent  road  cutting,  East  Ogwell,  near  Newton  Abbot,  south  Devon.  Also 
Frasnian  of  Germany  (Harz)  and  Poland  (Kielce). 

Description.  Nothing  is  known  of  colony  shape,  size  and  external  features  from 
the  present  material. 

The  colonies  are  pseudocerioid,  tending  to  astraeoid  and  occasionally  thamnasterioid 
in  parts.  Corallites  are  irregularly  polygonal,  usually  pentagonal  or  hexagonal,  and 
for  the  most  part  separated  from  each  other  by  a  straight  or  slightly  zigzagged  wall 
formed  by  the  deflection  of  the  peripheral  septal  ends.  When  this  pseudotheca 
breaks  down,  the  geniculate  ends  of  the  septa  are  more  or  less  confluent  with  those 
of  the  adjacent  corallite. 

The  septa,  of  two  orders  major  and  minor,  are  variable  in  thickness  but  average 
about  0-05  mm.  in  the  dissepimentarium  where  they  may  be  sinuous.    Occasionally 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  243 

corallites  have  a  bilateral  appearance  when  the  septa  tend  to  lie  parallel  on  either 
side  of  the  tabularium.  Within  the  tabularium,  the  major  septa  thin  considerably 
to  about  0-02  mm.  in  thickness  and  continue,  as  straight  or  flexuous  filaments,  more 
or  less  to  the  axis.  Minor  septa  are  confined  to  the  dissepimentarium.  Sometimes 
the  axial  ends  of  adjacent  major  septa  fuse,  or  the  axial  ends  of  opposite  septa  are 
continuous  across  the  tabularium.  Signs  of  thickening  of  the  axial  ends  of  septa 
are  very  rare. 

The  septa  are  variably  carinate  in  the  dissepimentarium.  In  exceptional  cases 
the  carinae  may  reach  0-4  mm.  in  width,  but  mostly  they  are  between  o-r  and  0-15 
mm.  Their  spacing  varies  but  normally  there  are  10  carinae  per  mm.  of  septal 
length.  On  individual  septa,  carination  is  heaviest  in  a  zone  of  irregular  width 
surrounding  the  tabularium,  corresponding  to  the  septal  dilatation  observed  in 
Frechastraea  pentagona  and  F.  goldfussi;  occasionally,  adjacent  carinae  may  fuse 
along  the  septa  in  this  zone.  As  they  enter  the  tabularium,  the  septa  rapidly  become 
smooth. 

The  dissepiments,  moderately  to  closely  spaced  in  cross-section,  are  almost 
exclusively  uniserial  between  adjacent  septa.  The  tabularium  junction  is  sharply 
defined. 

In  longitudinal-section  the  dissepimentarium  consists  of  several  series  of  regularly 
developed,  small,  arched  dissepiments.  Their  height  is  usually  between  o-i  and 
o-2  mm.  The  dissepiments  become  progressively  more  globose  towards  the  tabular- 
ium and  against  the  tabularium  boundary,  horseshoe  dissepiments  occur  very  rarely. 
The  surface  of  the  dissepimentarium  slopes  slightly  downwards  and  outwards  from 
the  tabularium  and  is  flat  lying  peripherally. 

The  tabularium  structure  is  usually  simple.  The  tabulae  are  commonly  incomplete 
and  may  be  flat  or  moderately  arched.  Their  vertical  spacing  averages  about 
o-2  mm.  In  one  specimen,  OUM  D3iob,  there  is  evidence  of  highly  domed  axial 
plates,  occupying  about  one  fifth  of  the  tabularium  diameter,  with  peripheral  flat, 
or  slightly  bowed  tabulae.  This  axial  structure  is  not  clearly  developed  but  it 
appears  to  be  similar  to  that  described  in  F.  goldfussi. 

Very  few  instances  of  increase  have  been  observed  but  both  axial  and  lateral  are 
represented. 

A  total  sample  ot  ten  colonies  from  Ramsleigh  Quarry  and  the  adjacent  road 
cutting  has  been  statistically  analysed.  Four  of  the  colonies  have  also  received 
individual  treatment.  The  statistics  are  listed  in  Table  9  and  illustrated  graphically 
in  Text-figs.  15,  16. 

Discussion.  Material  assigned  to  this  new  species  was  formerly  known,  due  to 
Freeh's  (1885  :  56)  work,  as  '  Phillip sastrea  pentagona  var.  micrommata  (C.  F. 
Roemer)  '.  Recently,  however,  the  writer  has  been  able  to  section  Roemer's 
figured  specimen  of  Smithia  micrommata,  which  is  described  above.  This  shows 
conclusively  that  Freeh  and  all  subsequent  workers  have  wrongly  interpreted 
Roemer's  species. 

Freeh  also  placed  Acervularia  roemeri  var.  /?  concinna  F.  A.  Roemer  in  synonymy 
with  his  P.  pentagona  var.  micrommata.  The  original  material  of  the  former  variety 
appears  to  be  missing  and  F.  A.  Roemer's  illustrations  (1855,  pi.  6,  figs,  iga-c)  are 


244  COLONIAL  PHILLI  PS  ASTR  AEI  D  AE  FROM  S.E.  DEVON 

40 


f     20 


f      20 


olony  2 


10 


%t 


i  t 


0  8  10  1-2 

dt       (mm) 


1-4 


12 
11 
10 

9     • 

8 

7 


08  10 

dt       (mm) 


i  * 


1-2 


1-4 


Total  sample  Colony        1  Colony       2  Colony         3  Colony       4 

Fig.  15.     Frechastraea  carinata. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


245 


rather  poor.  They  show  no  sign  of  septal  carinae,  however,  and  the  septa  themselves 
are  spindle-shaped  and  thin  peripherally — the  phillipsastreid  rather  than  the 
frechastraeid  pattern.  The  present  material  is  thus  considered  distinct  from  A. 
roemeri  var.  /?  concinna. 

F.  carinata  is  extremely  close  quantitatively  to  F.  pentagona,  with  its  mean  tabular- 
ium  diameter  falling  between  that  of  F.  ■pentagona  pentagona  and  F.  pentagona 
minima  (Text-fig.  16).     Statistical  discrimination  however  reflects  the  high  sample 


Table  9. — Statistical  data  for  some  characters  of  Fr&chaslraea  carinata. 


Total  sample        Colony  1  Colony  2  Colony  3  Colony  4 


N 


423  (9) 


60 


60 


60 


60 


O.R.            ( 

3-8-1-3 

0 ■ 8-1 • 0 

O-8-I  -2 

0 • 8-1 • 2 

] 

[-0-1-3 

X 

I  -02 

o-93 

0-96 

I  -02 

1-16 

dt 

s 

O-II 

0-057 

0-081 

0-069 

0-064 

C.V. 

10-56 

6-i8 

8-44 

6-8i 

5-47 

S.E.m 

0-005: 

2 

0-0074 

O-OIO 

0 • 0090 

00082 

O.R 

8-I4 

8-10 

8-1 1 

8-1 1 

9-1 1 

X 

9-75 

9-42 

9-35 

9-87 

9  95 

n 

s 

o-34 

029 

0-40 

0-19 

0-I2 

C.V. 

3-5i 

3-°7 

4-29 

1-89 

1-18 

S.E.m 

0-017 

0-037 

0-052 

0024 

0-015 

O.R.           7 

•27-12-' 

73 

8-89-11-2 

6 

8- 00-11 -ii 

8.33-12-50 

7 

■50-12-50 

X 

9-66 

10-20 

9-79 

9-74 

8-57 

n/dt 

s 

o-68 

0-32 

0-41 

0-58 

o-39 

C.V. 

7-o7 

3-13 

4-22 

5-9o 

4-5o 

S.E.m 

0-033 

0-041 

0-053 

0-074 

0-050 

Graphs: — 

r 

o-95 

I-OO 

o-93 

0-70 

0-89 

n/dt 

a 

3-i8 

5-07 

4-96 

2-69 

1-84 

b 

6-52 

4-73 

4-60 

7-13 

7-82 

n/dt/ 
/dt 

r 
a 

—  i- 00 
-6-36 

—  I  -oo 
-5-59 

—  096 

—  0-96 
-8-30 

—  I -00 

-6-o6 

/ 

b 

16- 12 

15-37 

14-69 

18-18 

15-62 

A 

At 

At/A 

Graph : — At/ A 

O.R. 

o-io-o- 17 

0-0067- 

-o-oii         o- 

052- 

-0  •  088 

X 

0-13 

0 

0082 

0 

■063 

r 

0-77 

s 

0-024 

0 

0013 

0 

•on 

a 

0-054 

C.V. 

18-45 

15 

■86 

17 

•68 

b 

0-0012 

S.E.m 

00081 

0 

•00044 

0 

■0037 

GEOL.  15, 

5- 

25 

246 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 
500    r 


400 


300 


200 


100 


06 


12  14 

dt       (mm) 


20  2-2 


•030    r 


•02  5      ■ 


At 

(sq.  cms) 


020 


015 


010 


005 


1^-f 


^.C^^^X 


010  0  14 


018 

A 


13   r 
12    • 
11 
10 
9 


0-22  026 

(sq.  cms) 


12 


0-30  034 


038 


10    • 


/& 


6    • 


06  10  14  18  22 

dt      (mm) 


06  10  14 

ut      (mm) 


1-8  22 


f.   pentagona  pentagona      — —      (")  f.   pentagona  minima      —  ———  —  —     (♦) 

f.    carinafa (°)  f    goldfussi       ^^^-^^^—      (•)  f.    bowerbanki      (•) 

Fig.  16.     Graphical  comparison  of  some  quantitative  characters  for  the  species  and  sub- 
species of  Frechastraea. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


247 


Table  10. — Statistical  discrimination  between  some  characters  for  species  and  subspecies  of 
Frechastraea.     Significant  values  in  bold;  s  =  slope,  p  =  position. 

pentagona    pentagona  .  ,,,      .     .         ,      ,. 

.      .  .    .  carinata      goldfussi     bowerbanhi 

pentagona      minima 


0-149 


Values 

oft":— 

pentagona  pentagona 

15-388 

9-838 

55-924 

pentagona  minima 

15-388 

6-589 

dt              carinata 

9-838 

6-589 

goldfussi 

55-924 

bowerbanki 

0-149 

Values  of  "z": — 

pentagona  pentagona 

s 
P 

11  041 

0-724 
6-946 

pentagona  minima 

s 

p 

11  041 

11-335 

n/dt    < 

carinata 
goldfussi 
bowerbanki 

s 

p 

s 

p 

s 

p 

0-724 
6-946 

11-335 

1-562 
10  426 

1  -562 
10  426 

13-855 

pentagona  pentagona 

s 

p 

1-326 
0-116 

o-379 
4  533 

2-196 
3-724 

pentagona  minima 

s 

p 

1-326 
0-116 

1-544 
o-ooo 

0-719 
3-496 

At/A  - 

carinata 

s 

p 

o-379 
4  533 

1-544 
o-ooo 

goldfussi 

s 

p 

2-196 
3-724 

0-719 

3-496 

bowerbanki 

s 

p 

3  719 

13-855 


3-719 


sizes  and  F.  carinata  can  be  shown  to  be  significantly  different  from  both  subspecies 
in  dt  and  the  plot  of  n  against  dt,  and  from  F.  pentagona  pentagona  alone  in  the 
plot  of  At  against  A  (Table  10). 

Qualitatively,  F.  carinata  is  distinguished  by  its  sinuous,  variably  carinate  septa. 
There  is  also  a  characteristic  tendency  for  the  majority  of  the  septa  in  one  corallite 
to  follow  the  same  directional  trend  in  the  dissepimentarium.  F.  sanctacrucensis 
(Rozkowska),  recorded  so  far  only  from  Poland,  is  closest  in  general  appearance  to 
F.  carinata.  The  former  is  clearly  astraeoid,  however,  with  considerably  larger 
tabularia  (dt  ca.  2  mm.,  n  12-14),  and  its  septal  carination  gives  a  distinctive  "  string- 
of-pearls  "  effect. 


Frechastraea  goldfussi  (de  Verneuil  &  Haime) 
Plate  10,  figs.  1-5;   Plate  11,  figs.  1,  2 

1826     Cyathophyllum  ananas  Goldfuss  :  60  pars,  pi.  19,  fig.  4a  (non  fig.  46). 

1840     Astrea  (Favastrea)  pentagona  (Goldfuss)  Lonsdale  :  697  pars,  pi.  58,  fig.  la  (non  fig.  1). 

1850     A cervularia  goldfussi  de  Verneuil  &  Haime  :  161. 


248  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

1851  Acervularia  goldfussi  de  Verneuil  &  Haime;    Edwards  &  Haime  :  417. 

1851  Acervularia  limitata  Edwards  &  Haime  :  419. 

1853  Acervularia  goldfussi  de  Verneuil  &  Haime;   Edwards  &  Haime  :  236,  pi.  53,  figs.  3,  3a. 

1853  Acervularia  limitata  Edwards  &  Haime;    Edwards  &  Haime  :  238,  pi.  54,  ?figs.  1,  \a. 

1881  Acervularia  pentagona  (Goldfuss) ;    Schluter  :  89,  pi.  9,  figs.  4,  5. 

1883  Acervularia  goldfussi  de  Verneuil  &  Haime;    C.  F.  Roemer  :  352. 

1883  Acervularia  limitata  Edwards  &  Haime;    C.  F.  Roemer  :  353. 

1885  Phillipsastrea  ananas  (Goldfuss)  Freeh  :  49  pars. 

1885  Phillipsastrea  pentagona  (Goldfuss)  Freeh  :  54,  pars,  pi.  3,  ?figs.  6,  9. 

1951  Phillipsastraea  limitata  (Edwards  &  Haime);    Soshkina  :  97,  pi.  17,  fig.  2;    pi.  18,  fig.  2; 
pi.  23,  fig.  4. 

1952  Phillipsastraea  limitata  (Edwards  &  Haime);    Soshkina  :  101,  pi.  42,  fig.  142. 

1953  Phillipsastraea  goldfussi  (Edwards  &  Haime) ;    Rozkowska  :  62,  text-figs.  35-37,  pi.  8, 
figs.  5.  6. 

1958  Billingsastraea  goldfussi  (Edwards  &  Haime)  Schouppe  :  236,  text-figs.  25,  26. 

1959  Phillipsastraea  goldfussi  (Edwards  &  Haime);   Middleton  :  156,  text-fig.  6d. 

But  not: 

1881     Heliophyllum  cf.  limitatum  (Edwards  &  Haime)  Schluter  :  87,  pi.  8,  figs.  1,  2. 

Diagnosis.  Pseudocerioid  Frechastraea.  Mean  tabularium  diameter  1*53  mm. 
and  between  7  and  17  major  septa  (East  Ogwell  sample).  Septa  smooth,  very  rarely 
carinate,  usually  with  lobate  thickening  on  axial  ends  of  major  septa.  Dissepi- 
mentarium  occasionally  with  imperfect  series  of  horseshoe  dissepiments  at  tabularium 
boundary.  Tabularium  composed  of  complete  or  incomplete  tabulae,  rarely  with 
axial  domes  of  horseshoe  section.     Increase  axial  or  lateral. 

Holotype.  (see  de  Verneuil  &  Haime,  1850  :  161).  The  original  of  Goldfuss' 
(1826,  pi.  19,  fig.  4«)  illustration  of  Cyathophyllum  ananas.  Frasnian;  Namur, 
Belgium.     This  specimen  is  either  mislaid  or  lost. 

Material.  Ramsleigh  Quarry:  TM(JB)3io  (Colony  1),  BM(NH)  R46370 
(Colony  2),  TM(JB)3o6  (Colony  3),  BM(NH)  R23208  (Colony  4),  BM(NH)  R46369 
(Colony  5),  TM(JB)307  (Colony  6),  BM(NH)  R23217  (Colony  7),  BM(NH)  R46367 
(Colony  8),  BM(NH)  R46368  (Colony  9),  BM(NH)  R23302  (Colony  10).  Other 
measured  specimens:  TM(JB)305B,  TM(JB)3ir-3i3,  TM(JB)3i8,  BM(NH)  R46374, 
BM(NH)  R5636,  BM(NH)  R5642,  BM(NH)  R5648,  BM(NH)  R23301,  OUM  D530-1, 
OUM  D539-41. 

South  Devon:   GSM  (Geol.  Soc.  Coll.)  6183. 

Distribution.  England:  Lower  Frasnian  limestones,  Ramsleigh  quarry,  near 
Newton  Abbot,  south  Devon.  Also  Frasnian  of  Belgium,  Germany,  Poland  and 
U.S.S.R.  (Timan) ;  ?  Frasnian  of  Spain. 

Description.  The  specimens  are  incomplete  colonies  frequently  comprising 
more  than  200  corallites.  Colony  shape,  size  and  external  features  are  unknown 
from  the  present  material  but  according  to  Rozkowska  (1953  :  62)  Polish  representa- 
tives of  the  species  are  hemispherical  or  lenticular  colonies  up  to  7-5  cm.  diameter 
and  3*5  cm.  in  height,  covered  basally  by  a  holotheca.  Where  seen  in  the  English 
material,  the  holotheca  is  about  o*i  mm.  thick  (PI.  10,  fig.  3).  The  calices,  after 
Rozkowska,  are  deep  with  a  flat  floor,  surrounded  by  an  annular  rim  at  the  tabularium 
boundary. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  249 

The  colonies  are  massive,  pseudocerioid.  The  corallites,  usually  pentagonal  or 
hexagonal,  are  separated  from  each  other  by  a  strong,  straight  or  more  frequently 
zigzagged  pseudotheca.  Infrequently  the  pseudotheca  may  break  down  in  part 
when  the  septa  are  more  or  less  confluent  from  one  corallite  to  the  next. 

In  the  dissepimentarium  the  septa,  major  and  minor,  are  usually  about  0-05  mm. 
but  may  occasionally  reach  0-2  mm.  in  thickness.  They  are  straight  and  are  variably 
dilated  for  a  short  length  against  the  tabularium  boundary.  The  major  septa 
continue  as  extremely  thin  processes  across  the  tabularium  to  the  axis  where  they 
develop  a  variable  lobate  thickening.  Usually  these  septa  are  fractionally  with- 
drawn from  the  axis  when  their  dilated  ends  form  a  pseudoaulos.  The  minor  septa, 
normally  less  dilated  than  the  major,  are  not  continued  beyond  the  thickened  portion 
which  may  or  may  not  project  slightly  into  the  tabularium.  The  septa  are  usually 
smooth  sided  but  may  occasionally  be  lightly  carinate. 

The  dissepiments  appear  fairly  widely  spaced  in  cross-section  and  are  almost 
always  uniserial  between  adjacent  septa.  Occasionally  the  traces  of  horseshoe 
dissepiments  can  be  distinguished  around  the  periphery  of  the  tabularium,  corre- 
sponding to  the  zone  of  septal  dilatation.  The  tabularium  junction  itself  is  sharply 
denned. 

In  longitudinal-section  the  dissepimentarium  is  composed  of  several  series  of 
small,  well  arched  dissepiments.  The  number  of  series  is  variable,  however,  and 
may  rarely  be  as  low  as  two.  Usually  the  dissepiments  are  regularly  developed  with 
a  height  of  about  0-2  mm.  and  become  somewhat  more  globose  at  the  boundary  with 
the  tabularium.  Occasionally,  however,  the  series  immediately  adjacent  to  the 
tabularium  may  become  modified  in  part  to  form  an  incomplete  and  irregular  series  of 
horseshoe  dissepiments.  The  dissepimental  surface  slopes  away  from  the  tabularium 
for  a  short  distance  and  is  flat  lying  peripherally. 

The  tabularium  structure  is  usually  simple,  varying  from  complete  flat  to  slightly 
domed  tabulae,  to  wide  slightly  arched  incomplete  vesicular  tabulae.  The  vertical 
spacing  of  the  plates  varies  between  o-i  mm.  and  0-5  mm.  but  is  usually  about  0-2 
mm.  Occasionally  steep  sided  complete  tabulae  with  narrow  flat  crests  are  developed 
and  rarely  highly  globose  vesicles  with  a  horseshoe-shaped  section  appear  in  the  axis 
of  the  tabularium.  When  two  or  three  of  the  latter  vesicles  are  superposed,  peripheral 
plates  slope  steeply  downwards  and  outwards  from  them. 

Increase  is  axial  or  lateral,  the  latter  occurring  more  commonly. 

A  statistical  analysis  has  been  made  of  specimens  of  this  species  from  Ramsleigh 
Quarry.  The  total  sample  comprises  25  colonies,  10  of  which  have  been  analysed 
individually.  The  statistics  are  given  in  Table  ri  and  the  data  is  presented 
graphically  in  Text-figs.  16-18. 

Discussion.  This  species  was  erected  by  de  Verneuil  &  Haime  (1850  :  161)  in  a 
fossil  list,  but  as  they  stated  that  their  new  species  was  Cyathophyttum  ananas 
Goldfuss  (1826,  pi.  19,  fig.  4«,  non  fig.  46)  their  designation  is  valid.  Acervularia 
goldfussi  was  later  described  in  some  detail  by  Edwards  &  Haime  (1851,  1853)  and 
many  subsequent  workers  have  mistakenly  attributed  the  species  to  these  authors. 

The  lectotype  of  Acervularia  limitata  Edwards  &  Haime  selected  by  Soshkina 
(1951  :  97)  is  lost  and  the  figure  of  the  specimen  (Edwards  &  Haime  1853,  pi.  54, 


25° 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


f     20 


f     20 


f      20      • 


f     20 


f      20       • 


20         2-2 


1-8  20         22 


Fig. 


—  —  —  —  —  —  —       Total    sample 

17.     Tabularium  diameter  frequency  curves  for  ten  colonies  of  Frechastraea  goldfussi. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  251 


10 


8     • 


% 


6      • 


10 


12 


16  1 

at        (mm) 


20  2-2 


T  Total   sample 

Fig.  18.     Graphs  of  septal  insertion  and  septal-tabularium  ratio:  tabularium  diameter  for 
ten  colonies  of  Frechastraea  goldfussi. 


25-2 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


t 


>>  o 
a    n 
o 
"o 
U 


>>  o 
a  vo 
o 


>-.     ro 


o 

O 


a 

ro 

T3 

>>  vO 

rt 

g     vo 

u 

en 

O 

u 

>, 

N 

l4 

c 

00 

H 

o 

"o 

w 

O 

1-1 

M 

< 

H 

H 

>, 

o\ 

c 

>n 

_o 

o 

O 

__ 

* 

In 

nS 

"ft 

0 

H 

a 

rd 

O 

o 

m 

•   <0  oo 

0>  00    o 

N  M     O 


o 

ro  O 


O    m    O    N    O 


O       00        o\ 

O^  Nli^O 
fl   NO    MO 

N 

"  O   to  m  O 

IT)                           ^ 

.    ro  n  ro  n 

(^  IT,   H     U1   0 

NfOfl 
00  00  00 

00    w 

*0    M     O     Tf  0 
M 

O   h   o   M   0 

M     M 

0\*O    O    N    0 
N 

O  n  + 

O    <N 

1      1 

o 

W 

1 

'     -3-vO  M3    O 

<1> 

o 
ca 

a 

_o 
"o 

O 

m   u-)  o    ■*■  O 

tHO+o 

in 

O 

XS 

o 

<D 

Th 

n     oo       <o 

6 

•     00     Q\   Q\   M 

>> 

0 

>h    ■+  O    "-)  0 

o 

a 

o 

w 

o 

N    H    o  O    O 

m 

«   O10    ft 
M    N  to  N    O 

Ch    O   rtO 


n  o  ^j-  a»  h 

H     C>    M     M     0 


■*■  ro  -J-  ^-  "J- 

m  00    ■*■  N  0 


to  n  o  o  N 

w  00    N    N  0 


ntonao 
iso  to 


•     C\  in  ^-  ro 

oo    "">  N  N    0 

to    NO  f^O 
00 

o 
N 

vo  ^ 

.    n  m  in  n 

00  O    <N  N  O 

in  n  o  n  o 


.  m  in  N  ro 
00  N  N  N  O 
I      ■ 


m    O    "    0         N  N  O    ro  0 


00    <3\  O    ro  o  OvvO    O   fOO 


•      Ol  +N.Tf 
Q\   TT  CO  IT)  o 


N    M    0  00    O 


VO  M  0 

•       OV    O    M     M 

m    co  O    "">  O 


On  O   O    •*•  O 


O 
ro  O  u~>  u-jvo 
m    ■<*•  u-l  IN    O 

oi  o  o  io  o 


M      N    O      Tf    O 


N  N  0   ro  0 


rorOMM  ro  N    rt-  Ov  in 

fO  m    0    O  m    CO  Tt-00    O 

CAhO    BIO 


M     M     O     O     0 


•     rOiO   O    0         NO   w   N  w 

NinMMO       MMinino 


S 

&        >  w 

0 IX  <e  d  05 


s 
Pi       >w 

0  IX  to  U  05 


,  to  'i-oo  0 

NO    iflO 


e 

O  ix  oo  d  CO 


N  N     Tt- 

o  N  oo 


o  ro  N 


OO      M      f^ 

ovoo  in 

O    ro  ro 

I  I  H 


I  I  " 


0    N    N        O    ro  ro 


O    N 

I      I 


I  I 


I  I 


I  I 


0   roo         O   N 


u  a  A        S.   "5  J2 


J3  \ 


J3 

rt 


1-    S£ 


ro 


"   O   No 


ro  O    0    N  o 

N  M 

o 
o 


00 
N  Ti- 

O   Oi"l 

•       M     O 

o  o  o 

■*■  o 


00 


O    N  o 


N  O 

■     O^  rOvO   O 
O    m    o    "J-  o 

o  o  o  «io 


o  ix  w  d  «5 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  253 

figs.  I,  la)  is  of  uncertain  affinity.  The  only  survivor  of  the  original  syntypes  of 
Acervularia  limitata  appears  to  be  the  specimen  figured  by  Lonsdale  (1840  pi.  58, 
fig.  la,  non  fig.  1)  as  Astrea  (Favastrea)  pentagona  which  is  GSM  (Geol.  Soc.  Coll.) 
6183.     This  specimen  is  conspecific  with  FrecJiastraea  goldfussi. 

Freeh  (1885  :  49,  footnote)  considered  Acervularia  goldfussi  as  interpreted  by 
Edwards  &  Haime  (1851,  1853)  to  be  transitional  in  form  between  his  Phillipsastrea 
ananas  and  '  P.  '  pentagona  and  accordingly  split  the  former  species  between  the 
latter  two.  This  explains  the  apparent  inconsistencies  in  Freeh's  synonymies. 
Under  P.  ananas  (p.  49)  he  placed  Cyathophyllum  ananas  Goldfuss  (1826  pi.  19, 
figs.  4a,  b)  whilst  under  P.  pentagona  he  listed  '  Acervularia  goldfussi  de  Verneuil  & 
Haime  1850  p.  161  e.p.  '.  Freeh  considered  Edwards  &  Haime's  (1853,  pi.  53, 
figs.  3,  3«)  figured  specimen  of  Acervularia  goldfussi  as  conspecific  with  hisP.  pentagona. 

Most  subsequent  authors  appear  to  have  interpreted  F.  goldfussi  on  Edwards  & 
Haime's  figures  and  descriptions  and  to  have  considered  it  worthy  of  specific  rank, 
although  closely  related  to  F.  pentagona.  In  the  absence  of  Goldfuss'  type  specimen, 
this  interpretation  is  followed  here.  The  similarity  between  F.  goldfussi  and  F. 
pentagona  pentagona  is  considerable,  both  being  pseudocerioid  with  essentially 
straight,  non-carinate  septa.  Differences  are  confined  to  small  details  such  as  the 
lobate  thickenings  on  the  axial  ends  of  the  major  septa  in  F.  goldfussi  and  the  higher 
incidence  of  horseshoe  dissepiments  in  this  species.  F.  goldfussi  and  F.  pentagona 
pentagona  are,  however,  clearly  differentiated  on  quantitative  characters  (Text-fig.  16 
and  Table  10).  The  former  has  a  considerably  larger  tabularium  diameter  than  the 
latter  and  the  two  are  significantly  different  in  their  growth  lines  on  graphs  plotting 
n  against  dt  and  At  against  A.  Both  differ  strongly  from  Phillipsastrea  ananas 
(described  on  p.  228)  through  the  larger  dimensions,  the  spindle-shaped  septa, 
complex  tabularium  and  row  of  well  developed  horseshoe  dissepiments  of  the  latter. 

Rozkowska  (1953  :  62  et  seq.)  has  described  '  Phillipsastraea  '  goldfussi  in  detail 
from  the  Upper  Frasnian  of  Poland.  Her  material  has  a  greater  range  (1-2-2-8  mm.) 
and  a  higher  mean  value  (i-8  mm.)  for  the  tabularium  diameter  than  the  English 
representatives.  There  is  no  doubt  that  the  two  samples  are  conspecific,  however, 
and  the  slight  size  difference  may  be  due  to  the  higher  stratigraphical  level  of  the 
Polish  collection. 

Attention  is  drawn  for  the  first  time  to  the  rare  horseshoe  dissepiments  developed 
in  representatives  of  this  species.  Besides  the  English  examples,  Schluter  (1881, 
pi.  9,  fig.  5)  illustrated  a  specimen  as  A.  pentagona  referable  to  this  species  from  the 
Frasnian  of  Stolberg,  near  Aachen  (Germany)  which  also  clearly  shows  occasional 
horseshoes  developed  against  the  tabularium  junction. 

Frechastraea  bowerbanki  (Edwards  &  Haime) 
Plate  11,  Fig.  3;   Plate  12,  Figs.  1-3 

1851  Smithia  bowerbanki  Edwards  &  Haime  :  423. 

1852  Acervularia  seriaca  Quenstedt  :  664,  pi.  60,  fig.  3. 

1853  Smithia  bowerbanki  Edwards  &  Haime;    Edwards  &  Haime  :  241,  pi.  55,  figs.  2,  2a. 
1879  Acervularia  seriaca  Quenstedt;   Quenstedt  :  536,  pi.  163,  fig.  1. 


254  COLONIAL  PHILLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

1883     Phillipsastraea  bowerbanki  (Edwards  &  Haime)  C.  F.  Roemer  :  391. 
1885     Phillipsastraea  bowerbanki  (Edwards  &  Haime);    Freeh  :  63,  pi.  4,  figs,  g,  ga,  b. 
?i95i     Pachyphyllum  bowerbanki  (Edwards  &  Haime)  Soshkina  :  89,  pi.  16,  figs.  1,  2. 

But  not: 
1879     Smithia  bowerbanki  Edwards  &  Haime;    Quenstedt  :  536,  pi.  162,  fig.  39. 
1953     Phillipsastraea  bowerbanki  (Edwards  &  Haime) ;   R6zkowska  :  67,  pi.  8,  figs.  3,  4. 

Diagnosis.  Thamnasterioid  Frechastraea.  Mean  tabularium  diameter  1-54  mm. 
with  7  to  12  major  septa  (East  Ogwell  sample).  Septa  rarely  slightly  dilated  at 
tabularium  boundary.  Dissepiments  characteristically  weakly  arched  and  rather 
elongate;   flattened  horseshoes  extremely  rare.     Tabularium  structure  simple. 

Lectotype.  Selected  by  Soshkina  (1951  :  89).  The  original  of  Edwards  & 
Haime's  (1853,  pi.  55),  figs.  2,  2a.  Devonian;  Torquay,  south  Devon.  This  speci- 
men appears  to  be  lost. 

Material.  Ramsleigh  Quarry:  TM(JB)2g4a  (Colony  1),  TM136/7  (Colony  2). 
Other  measured  specimens:   BM(NH)  R46372-73. 

Distribution.  England:  Lower  Frasnian  limestones,  Ramsleigh  Quarry,  East 
Ogwell,  near  Newton  Abbot,  south  Devon.  Also  Frasnian  of  Germany  (Harz)  and 
?  the  U.S.S.R.  (southern  Urals). 

Description.  The  material  is  fragmentary  and  nothing  is  known  of  the  external 
features,  shape  and  overall  size  of  the  corallum. 

The  colony  is  thamnasterioid,  with  the  septa  of  adjacent  corallites  most  commonly 
perfectly  confluent  and  less  frequently  irregularly  abutting.  Occasional  septa  may 
have  free  ends  in  the  dissepimentarium.  The  tabularia  are  regularly  and  fairly 
widely  spaced. 

The  septa  are  o-05-o-i  mm.  thick  and  are  regularly  developed  between  tabularia. 
They  may  be  straight,  but  are  usually  sinuous  and  occasionally  geniculate.  In  some 
cases,  the  sides  of  the  septa  are  slightly  corrugated,  presumably  by  the  presence  of 
slightly  swollen  trabeculae;  they  are  never  truly  carinate  and  usually  the  septa  are 
smooth  sided.  Septal  dilatation  at  the  tabularium  boundary  is  slight  or  lacking. 
Usually  the  major  septa  project  into  the  tabularium  for  between  o-i  and  0-2  mm. 
with  no  change  in  thickness.  At  this  point,  they  thin  abruptly  and  continue  towards 
the  axis  as  strongly  attenuate  filaments.  Occasionally  the  major  septa  reach  the 
axis  but  more  commonly,  they  either  curve  sharply  to  become  confluent  with  an 
adjacent  or  nearby  septum,  or  they  end  about  a  third  of  the  tabularium  radius  short 
of  the  axis.     The  minor  septa  end  at  the  tabularium  junction. 

The  dissepiments  are  uniserial  between  adjacent  septa.  The  tabularium  junction 
is  strongly  and  sharply  defined. 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  several  series  of 
elongate,  weakly  arched  vesicles.  The  dissepiments  are  very  uniformly  developed 
with  a  height  of  o-i  mm.  The  surface  of  the  dissepimentarium  is  flat  peripherally, 
rising  slightly  to  a  crest  just  outside  the  tabularium  junction.  Rarely  a  flattened 
horseshoe  dissepiment  may  occur  among  the  series  forming  the  crest,  whilst  on  the 
tabularium  side,  a  vertically  discontinuous  row  of  normal  dissepiments  slopes  steeply 
axially  and  downwards. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  255 

50    r 


11    r 


10 


% 


1-2.  14  16  1-8 

dt     (mm) 


12  14  1-6  18 

at    (mm) 


Total   sample 


Colony       1 

Fig.  19.     Frechastraea  bowerbanki. 


Colony      2 


256 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


Table  12. — Statistical  data  for  some  characters  of  Frechastraea  bowerbanki. 


Total  sample 

Colony  1 

Colony  2 

N 

] 

'37  (4) 

48 

60 

O.R. 

1 

•3-1-8 

1 -3-i -8 

I-3-I-8 

X 

i-54 

i-54 

1-54 

dt 

s 

o- 12 

0-13 

o- 10 

C.V. 

7-73 

8-65 

6-56 

S.E.m 

O-OIO 

0-019 

0-013 

O.R. 

7-12 

8-1 1 

7-12 

X 

9-54 

9-65 

9-62 

n 

s 

0-25 

0-42 

o-37 

C.V. 

2-59 

4-3i 

3-89 

S.E.m 

0-021 

0060 

0-048 

O.R. 

4' 

71-7-69 

5-00-7-14 

5-00-7-14 

X 

6-24 

6-30 

6-36 

n/dt 

s 

o-34 

0-32 

0-26 

C.V. 

5-48 

5-06 

4  09 

S.E.m 

0-029 

0-046 

0-034 

Graphs  :- 

r 

0-89 

o-86 

o-86 

n/dt 

a 

2-08 

3-13 

3-70 

b 

6-34 

4-84 

3-91 

n/dt  . 

/ 

r 

-0-99 

-o-95 

-0-77 

/ 

dt 

a 

-2-88 

—  2-40 

-2-57 

/ 

b 

10 -66 

9-99 

10-32 

A 

At 

At/A 

Graph: — At/A 

O.R. 

o- 

29- 

-0-38        0 

•016- 

-0-020 

0-050-0-057 

X 

0 

'34 

0 

•018 

0-054 

r         0-96 

s 

0 

043 

0 

•0015 

00033 

a        0036 

C.V. 

12 

•56 

8 

•34 

613 

b        0-0062 

S.E.m 

0 

•021 

0 

•00077 

0-0017 

Tabularium  structure  is  simple  and  is  formed  with  both  complete  and  incomplete 
tabulae.  The  complete  plates  are  saucer-shaped,  whilst  the  incomplete  plates  are 
flat  to  slightly  arched  vesicles,  horizontally  disposed  in  the  centre  of  the  tabularium 
and  sloping  steeply  downwards  periaxially.  The  vertical  spacing  of  the  tabulae 
varies  considerably  between  0-05  and  0-3  mm. 

Increase  is  lateral,  with  the  daughter  corallites  developing  in  the  dissepimental 
tissue  equidistant  from  the  surrounding  adult  tabularia. 

Only  four  incomplete  colonies  from  Ramsleigh  Quarry  were  available  for  analysis, 
two  of  which  have  been  selected  for  individual  treatment.  The  statistics  are  listed 
in  Table  12  and  illustrated  graphically  in  Text-figs.  16  and  19. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  257 

Discussion.  Although  the  lectotype  is  lost,  there  can  be  no  doubt  from  Edwards 
&  Haime's  (1853  :  241,  pi.  55,  figs.  2,  2d)  description  and  figures,  of  the  interpretation 
of  this  species.  Figure  2«  shows  clearly  the  distinctive  characters  in  cross-section 
which  have  been  described  in  the  present  material.  Furthermore,  they  state 
(1853  :  242)  that  the  diameter  of  the  '  wall '  (=  tabularium  diameter)  is  about 
two-thirds  of  a  line,  which  is  approximately  1-4  mm.  This  agrees  with  the  present 
observations  but  contrasts  with  the  measurements  given  by  Rozkowska  (1953  :  67) 
for  her  Phillipsastraea  bowerbanki.  Rozkowska's  specimen  is,  in  fact,  distinct  from 
the  present  species  and  belongs  to  Frechastraea  pentagona  minima  (see  p.  240). 

Quenstedt  (1852  :  664,  pi.  60,  fig.  3)  erected  a  new  species,  Acervularia  seriaca, 
which  from  his  figure  and  description  appears  to  be  conspecific  with  F.  bowerbanki. 
Freeh  (1885  :  63)  was  of  the  same  opinion  and  placed  Quenstedt's  species  in  the 
synonymy  for  his  Phillipsastrea  bowerbanki.  Later  Quenstedt  (1879  :  536,  pi.  163, 
fig.  1)  refigured  Acervularia  seriaca  and  in  the  same  work  (p.  536,  pi.  162,  fig.  39) 
also  described  and  figured  '  Smithia  bowerbanki  Edwards  and  Haime  '.  On  the 
evidence  of  the  figures,  the  specimen  of  Acervularia  seriaca  appears,  as  before,  to 
belong  to  the  present  species,  whilst  his  Smithia  bowerbanki  is  probably  referable  to 
F.  pentagona  minima. 

F.  bowerbanki  differs  greatly  from  the  species  of  Frechastraea  described  above. 
This  species  has  tabularia  corresponding  in  size  to  those  of  F.  goldfussi  but  the  number 
of  septa  at  any  given  diameter  is  strikingly  lower  in  the  former.  F.  bowerbanki  also 
occupies  a  distinctive  position  on  the  graph  of  tabularium  area  plotted  against 
corallite  area,  reflecting  the  relatively  wide  spacing  of  the  tabularia  (Text-fig.  16, 
Table  10). 

F.  bowerbanki  is  further  distinguished  by  its  thamnasterioid  form  and  particularly 
by  weakly  arched,  rather  elongate  dissepiments  in  contrast  to  the  globose  dissepiments 
usually  found  in  species  of  Frechastraea. 

Genus  THAMNOPHYLLUM  Penecke  1894 

1894  Thamnophyllnm  Penecke  :  563. 

1909  Phacellophyllum  Giirich  :  102. 

1935  Disphyllum  {Phacellophyllum};  Lang  &  Smith  :  546. 

1935  Thamnophyllum ;  Lang  &  Smith  :  563. 

1939  Disphyllum  {Phacellophyllum} ;  Hill  :  224. 

1939  Thamnophyllum;  Hill  :  227. 
partim  1940  Thamnophyllum;  Hill  :  260. 

1940  Phacelophyllum;  Lang,  Smith  &  Thomas  :  98. 
1940     Thamnophyllum;  Lang,  Smith  &  Thomas  :  133. 
1949    Phacelophyllum;  Stumm  :  36. 

1949     Thamnophyllum;  Stumm  :  36. 
partim  1949     Macgeea  {Thamnophyllum) ;  Schouppe  :  100. 

1949  Thamnophyllum;  Soshkina  :  77. 

partim  1950    Phacellophyllum  {Phacellophyllum) ;  Wang  :  219. 

1950  Phacellophyllum  {Thamnophyllum);  Wang  :  219. 

1951  Disphyllum  {Phacellophyllum};  Taylor  :  185. 

1951  Thamnophyllum;  Soshkina  :  74. 

1952  Thamnophyllum;  Soshkina  :  85. 


258  COLONIAL  PH  ILLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

1953  Thamnophyllum;  Rozkowska  :  13. 
partim     1953     Macgeea;  Rozkowska  :  18. 

1954a  Phacellophyllum;  Hill  :  26. 

1954  Thamnophyllum;  Soshkina  :  65. 
1956     Thamnophyllum;  Rozkowska  :  304. 
1956a  Macgeea  {Thamnophyllum) ;  Fliigel  :  48. 

partim  1956a  Macgeea  (Macgeea);  Fliigel  :  53. 

1956b  Macgeea  (Thamnophyllum);  Fliigel  :  361. 
partim  1956b  Macgeea  (Macgeea);  Fliigel  :  361. 

1956     Thamnophyllum;  Hill  :  281. 

1956  Phacellophyllum;  Hill  :  282. 

1957  Thamnophyllum;  Rozkowska  :  83. 

1958  Macgeea  (Thamnophyllum);  Fliigel  :  625. 

1958  Macgeea  (Thamnophyllum);  Schouppe  :  226. 

1959  Thamnophyllum;  Fliigel  :  115. 
partim  1959    Phacellophyllum;  McLaren  :  28. 

i960  Thamnophyllum;  Rozkowska  :  44. 

partim  1963  Macgeea  (Macgeea) ;  Schouppe  &  Stacul  :  288. 

1963  Macgeea  (Thamnophyllum);  Schouppe  &  Stacul  :  291. 

1964  Thamnophyllum;  Webby  :  9. 

Diagnosis.  Dendroid  or  phaceloid  rugose  corals  with  axial  and  in  one  species, 
lateral  increase.  Septa  of  two  orders,  usually  spindle-shaped  in  dissepimentarium. 
Dissepimentarium  typically  with  outer  series  of  flat  dissepiments  and  inner  series  of 
horseshoe  dissepiments.  In  some  species,  dissepimental  structure  obscured  by  stereo- 
plasmic  thickening.  Tabulae  complete  or  incomplete,  with  periaxial  plates  variably 
developed. 

Type  species.  Selected  by  Lang  &  Smith  (1935  :  564)  and  see  Fliigel  (1958  :  625). 
Thamnophyllum  stachei  Penecke  1894  :  594,  pi.  8,  figs.  1-3;  pi.  11,  figs.  1,  2.  Emsian 
{barrandei-S>c]\ich.ttr\) ;  Marmorbruch  am  Graz,  Austria. 

Distribution.  Lower  to  Upper  Devonian  of  Europe,  Asia,  Australia  and  North 
America. 

Discussion.  Penecke  (1894  :  563)  described  four  species  of  Thamnophyllum,  three 
of  them  new  and  the  other  T.  trigeminum  (Quenstedt)  (see  T.  germanicum  germanicum 
nom.  nov.,  p.  260).  From  these,  Lang  &  Smith  (1935  :  564)  selected  T.  stachei  as 
type  species  and  redescribed  this  and  Penecke's  other  species.  They  removed  T. 
trigeminum  from  Thamnophyllum  and  assigned  it  to  Phacellophyllum  Giirich,  which 
they  regarded  as  a  genomorph  of  Disphyllum.  Giirich  (1909  :  102)  had  placed 
only  '  Phacellophyllum  caespitosum  Goldf.  '  in  his  new  genus.  As  Lang  &  Smith 
(1935  :  547)  pointed  out,  Gurich's  figures  (1909,  pi.  31,  figs.  5a,  b)  were  copied  from 
Schluter  (1881,  pi.  9,  figs.  6,  7)  which  fixes  the  type  species  of  Phacellophyllum  as 
Lithodendron  caespitosum  Goldfuss. 

Schouppe  (1949),  in  a  detailed  consideration  of  species  of  Thamnophyllum  and 
Phacellophyllum,  was  the  first  to  place  the  latter  in  synonymy  with  the  former. 
Furthermore,  he  made  Thamnophyllum  a  subgenus  of  Macgeea  on  the  basis  of  their 
similar  structural  plan,  separating  M.  (Thamnophyllum)  and  M.  (Macgeea)  principally 
on  their  growth  form. 

Both  Soshkina  (1951, 1952, 1954)  and  Rozkowska  (1953, 1956,  1957,  i960)  followed 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  259 

Schouppe  in  considering  Phacellophyllum  as  synonymous  with  Thamnophyllum ,  but 
neither  supported  the  subgeneric  relationship  with  Macgeea :  Rozkowska  (1953  :  18, 
1957  :  102)  listed  the  characters  by  which  Macgeea  and  Thamnophyllum  may  be 
distinguished.  On  the  other  hand,  Hill  (1954a,  1956)  and  McLaren  (1959)  retained 
Thamnophyllum  and  Phacellophyllum  as  separate  genera.  McLaren  (1959  :  28) 
pointed  out  that  T.  stachei,  type  species  of  Thamnophyllum,  is  imperfectly  known, 
whereas  Phacellophyllum  has  adequately  described  type  material.  From  Lang  & 
Smith's  (1935)  descriptions,  he  regarded  the  synonymy  of  the  two  genera  as  by  no 
means  certain. 

The  development  of  periaxial  tabulae  in  the  type  species  of  Phacellophyllum  has 
led  to  an  involved  and  lengthy  exchange  on  the  taxonomic  position  of  this  genus  in 
recent  years  between  Fliigel  and  Schouppe.  Fliigel  (1956a  :  53,  19656  :  361), 
describing  the  peripheral  plates  in  the  tabularium  of  Lithodendron  caespitosum  as  a 
third  zone  of  normal  dissepiments,  compared  the  structure  of  the  species  in  longitu- 
dinal-section with  that  developed  in  typical  forms  of  Macgeea.  On  this  basis,  he 
placed  Phacellophyllum  in  synonymy  with  Macgeea  sensu  stricto  and  retained  Thamno- 
phyllum as  a  subgenus  of  Macgeea  characterized  by  only  two  zones  of  dissepiments, 
flat  and  horseshoe.  Schouppe  (1958  :  220  et  seq.)  rejected  this  classification.  He 
considered  the  '  third  dissepimental  zone  '  to  constitute  part  of  the  tabularium 
structure  and  stated  (p.  227)  that  no  sharp  line  could  be  drawn  between  the  develop- 
ment of  periaxial  tabulae  in  Macgeea  and  Thamnophyllum.  Schouppe  thus  retained 
his  1949  classification,  placing  Phacellophyllum  in  Thamnophyllum,  which  he  separated 
subgenerically  from  Macgeea  on  the  basis  of  their  contrasting  growth  form  and  grade 
of  general  structural  complexity.  Fliigel  (1959  :  115,  footnote)  later  criticized 
Schouppe's  (1958)  subgeneric  diagnoses  as  insufficiently  differentiated.  Nevertheless, 
he  subscribed  to  a  very  similar  classification,  that  of  Rozkowska  (1957),  considering 
not  only  Phacellophyllum  to  be  synonymous  with  Thamnophyllum,  but  the  latter 
to  be  generically  distinct  from  Macgeea.  Finally,  Schouppe  &  Stacul  (1963)  published 
a  detailed  consideration  of  the  genera  involved,  in  which  they  returned  almost  exactly 
to  the  position  held  by  Fliigel  (1956).  They  wrote  (1963  :  285)  that  '  .  .  .  the  princi- 
pal stress  in  the  systematic  classification  should  be  laid  on  the  appearance  of  vesicular 
elements  in  the  peripheral  area  of  the  tabularium.  Consequently,  all  those  forms 
having  peripheral  vesicles  in  the  tabularium  (even  if  only  sporadic  .  .  .)  must  be 
placed  in  Macgeea  (Macgeea).  Forms  with,  on  the  other  hand,  simple,  peripheral, 
'  split-open  '  tabulae  or  additional  periaxial,  sloping,  plate-like  elements — without 
peripheral  vesicles — belong  to  Macgeea  (Thamnophyllum).'  Thus  they  placed 
Phacellophyllum  in  synonymy  with  Macgeea  (Macgeea). 

In  the  writer's  opinion,  Phacellophyllum  is  a  junior  synonym  of  Thamnophyllum 
and  the  latter  is  generically  distinct  from  Macgeea.  The  periaxial  elements,  so 
conspicuous  in  some  specimens  of  Lithodendron  caespitosum  may  show  considerable 
variation  in  their  development  within  the  species  as  a  whole  (see  p.  268).  Further- 
more, they  occur  to  a  greater  or  lesser  degree  in  many  other  species  of  Thamnophyllum. 
It  seems  unreasonable  to  suggest  a  morphological  and  genetic  distinction  between 
the  periaxial  elements  developed  in  Phacellophyllum  and  Thamnophyllum  at  what 
must  be  considered  an  arbitrary  level  in  their  degree  of  structural  complexity.     Such 


a6o  COLONIAL  PHI  LLIPS  ASTR  AEI  D  A  E  FROM  S.E.  DEVON 

a  variable  character  cannot  be  used  as  the  basis  of  a  subgeneric  division,  especially 
as  it  cuts  across  the  more  obvious  distinction  in  growth  form.  Species  of  Macgeea 
are  simple,  conical  corals  which  only  rarely  show  budding.  In  a  collection  of  234 
individuals  belonging  to  six  species  and  subspecies,  Rozkowska  (1957  :  118)  found 
only  four  budding  specimens.  Thamnophyllum  and  Phacellophyllum,  on  the  other 
hand,  are  phaceloid  or  dendroid  forms  with  cylindrical  corallites.  Commonly  in 
Macgeea,  the  horseshoe  dissepiments  are  less  regular  in  form  and  superposition  with 
a  generally  more  complex  tabularium  and  dissepimentarium  structure.  In  addition, 
Macgeea  usually  shows  distinct  bilateral  symmetry  in  the  adult  stage  in  contrast  to 
the  radial  symmetry  of  Thamnophyllum  and  Phacellophyllum. 

As  McLaren  (1959  :  28)  remarked,  existing  knowledge  of  T.  stachei  is  rather  imper- 
fect, but  the  writer  believes  enough  is  known  from  the  work  of  Lang  &  Smith  (1935  : 
581)  to  permit  an  opinion  on  the  taxonomic  status  of  Thamnophyllum.  T.  stachei, 
from  Penecke's  figures,  Lang  &  Smith's  description  and  their  specimens  (BM(NH) 
R30990-94),  shows  evidence  of  all  the  characteristic  features  described  in  better 
known  species  of  Thamnophyllum.  As  Rozkowska  (1957  :  101)  stated,  it  is  a  primi- 
tive early  member  of  the  genus  in  which  the  tabulae  are  very  simple  and  widely 
spaced.  Stereoplasmic  thickening,  which  in  T.  stachei  almost  completely  obscures 
the  dissepimental  structure,  is  also  a  variable  factor  in  other  species  of  Thamnophyllum. 

Thamnophyllum  germanicum  germanicum  nom.  nov. 

1894     Thamnophyllum  trigeminum  (Quenstedt)  Penecke  :  596,  pi.  8,  figs.  4-6. 

J959  Thamnophyllum  trigeminum  trigeminum  Penecke;  Fliigel  :  117  (see  for  extensive  syno- 
nymy). 

i960     Thamnophyllum  trigeminum  trigeminum  Penecke;    Rozkowska  :  53. 

1963  Macgeea  (Thamnophyllum)  trigemina  trigemina  (Penecke)  Schouppe  &  Stacul,  text-figs.  5, 
?I9- 

1963     Macgeea  [Macgeea)  sp.  Schouppe  &  Stacul,  text-fig.  18. 

But  not : 

1879     Cyathophyllum  caespitosum  trigemme  Quenstedt  :  518,  pi.  162,  figs.  5-8. 

Diagnosis.     See  Rozkowska  (1956  :  310). 

Lectotype.  Selected  by  Fliigel  (1959  :  118).  UPG  891  (Collections  of  the 
Palaontologisches  Institut,  Graz)  labelled  by  Penecke  in  1892  as  "  Fascicular ia 
trigemina".     Givetian;   (?)  Auberg,  near  Gerolstein,  Eifel,  Germany. 

Description.     See  Rozkowska  (1956  :  310). 

Discussion.  Cyathophyllum  caespitosum  trigemme  Quenstedt  has  been  shown  by 
Fliigel  (1959)  to  belong  to  Favistella  (Dendrostella).  Hitherto  it  had  been  regarded 
as  a  species  of  Thamnophyllum,  largely  due  to  Penecke  (1894  :  596)  misinterpreting 
Quenstedt's  description  and  illustrating  as  Thamnophyllum  trigeminum  (Quenstedt) 
a  new  and  generically  different  species.  Subsequent  workers  had  relied  mainly  on 
Penecke's  work  for  the  identification  of  Quenstedt's  subspecies. 

Fliigel  (1959  :  117)  thus  described  Penecke's  material  as  Thamnophyllum  trige- 
minum trigeminum  Penecke.  The  retention  of  Quenstedt's  specific  name  for  Penecke's 
misidentified  material,  however,  is  not  in  accordance  with  Article  49  of  the  I. C.Z.N. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  261 

(1964)  and  a  new  name  is  required.  With  the  permission  of  Professor  Fliigel, 
Thamnophyllum  trigeminum  trigeminum  Penecke  is  hereby  renamed  Thamnophyllum 
germanicum  germanicum  nom.  nov.  after  the  country  of  origin  of  Penecke's  specimens. 


Thamnophyllum  germanicum  schouppei  nom.  nov. 
Plate  13,  figs.  1-4;   Plate  14,  figs.  1-4 

1949  Macgeea  {Thamnophyllum)  caespitosa  (Goldfuss)  var.  minus  (F.  A.  Roemer)  Schouppe  :  152 
pars,  pi.  10,  figs.  21-24;  pl-  I3.  ngs-  73~75  inon  pl-  IZ>  ngs-  38-39;  pi.  13,  figs.  76-77; 
pi.  14,  figs,  iooa-c). 

1956&  Macgeea  {Thamnophyllum)  minima  Schouppe,  Fliigel  :  361. 

1965     Thamnophyllum  cf.  trigeminum  Penecke;    Scrutton  :  186. 

But  not: 

1855     Diphyphyllum  minus  F.  A.  Roemer  :  29,  pl.  6,  figs.  I2a-c. 

Diagnosis.  Thamnophyllum  with  mean  corallite  diameter  4*2  mm.  and  12  to  20 
major  septa  (topotype  sample).  Axial  increase  with  three  or  four  buds  lacking 
caenogenetic  tissue  in  axils  of  branches.  Tabulae  complete  or  incomplete  with 
periaxial  plates  sporadically  developed.     Skeletal  elements  generally  unthickened. 

Holotype.  See  Fliigel  (19566  :  361).  UGP327  (Collections  of  the  Palaonto- 
logisch  Institut,  Graz).  The  specimen  is  labelled  "  Middle  Devonian,  Torquay  " 
only,  but  it  comes  without  doubt  from  the  Givetian  limestones  in  Dyer's  Quarry. 

Material.  Dyer's  Quarry:  OUM  D506  (Colony  i),  OUM  D507  (Colony  2), 
OUM  D508  (Colony  3).  Other  measured  material:  OUM  D271,  OUM  D509,  OUM 
D511.     Additional  material:   OUM  D272,  OUM  0504-5,  OUM  D510. 

Distribution.     Type  locality  only. 

Description.  Colonies  are  phaceloid,  up  to  100  cm.  in  diameter,  consisting  of 
close-spaced,  sub-parallel,  cylindrical  corallites.  External  features  are  unknown  as 
recent  weathering  has  removed  the  epitheca  to  expose  the  peripheral  ends  of  the 
septa.     Calices  have  not  been  observed. 

In  cross-section  the  corallites  are  circular  to  sub-circular.  The  epitheca,  normally 
about  o-i  mm.  thick,  is  frequently  preserved  within  the  matrix.  The  septa,  of  two 
orders,  are  slightly  spindle-shaped  in  the  dissepimentarium.  The  minor  septa  may 
or  may  not  penetrate  very  slightly  into  the  tabularium  but  the  major  septa  usually 
reach  a  half  to  two-thirds  the  distance  to  the  axis.  The  septa  are  normally  straight 
in  the  dissepimentarium  but  the  major  septa  often  become  slightly  curved  or  sinuous 
in  the  tabularium. 

The  traces  of  the  sides  of  the  horseshoe  dissepiments  form  a  distinctive  double  wall 
in  cross-section.  Between  adjacent  septa,  the  two  walls  are  convex  towards  each 
other,  reflecting  the  saddle-shaped  form  of  the  horseshoe  dissepiments.  They  are, 
on  average,  about  0-4  mm.  apart,  with  the  outer  wall  the  same  distance  from  the 
epitheca.  The  innermost  wall  is  the  boundary  between  the  dissepimentarium  and 
the  tabularium. 

GEOL.  15,  5.  26 


262 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 
50 


40 


(mm) 


30      ■ 


20 


10 


235  3-15  395  4  75 

u     (04  mm   class    intervals) 


555 


100 

80 

60 

f 

40 

h  \  /\ 

20 

/  XV     \ 

0 

-r^-r*-— '■    '  i       i      i — ^r*-   |\   i       f*»  i 

30     ■ 

.-^ 

.^ 

^ 

n      20 

^ 

<? 

1 

23456789        10      11 

d        (1mm  class    intervals) 

T-      9 

schouppei                                T.      g      skalense 

T-      < 

I 

germonicum                             T.       g.      pajche/ae 

Total    sample  Colony       1  Colony     2  Colony     3 

Fig.  20.     Thamnophyllum  germanicum  schouppei.     Inset:    Graphical  comparison  of  some 
quantitative  characters  for  the  subspecies  of  Thamnophyllum  germanicum. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  263 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  an  outer  series  of 
flat  plates  and  an  inner  series  of  horseshoe  dissepiments.  The  outer  plates  are  not 
always  seen  as  they  are  easily  removed  by  erosion.  They  are  slightly  irregular  in 
spacing,  usually  between  0-2  and  0-5  mm.  apart  and  always  uniserial.  They  may 
be  perfectly  flat,  slightly  concave  or  convex.  The  horseshoe  dissepiments  are  thin 
walled  and  regularly  developed  in  a  single  series,  normally  between  three  and  four 
to  1  mm. 

The  tabulae  are  complete  or  incomplete,  often  with  irregularly  developed  periaxial 
plates.  They  may  be  flat  or  slightly  bowed  or  arched,  and  sometimes  strongly 
convex  towards  the  calice.  Tabulae  in  the  form  of  flat-topped  domes  are  rare. 
The  spacing  of  the  tabulae  is  variable  but  averages  about  18  to  1  cm. 

Increase  is  exclusively  axial  with  three  or  four  daughter  corallites  produced  in 
each  case.  There  is  no  indication  of  extra-dissepimental  tissue  (caenogenetic  tissue 
of  Soshkina  1953)  between  the  newly  formed  buds.  As  soon  as  normal  dissepimental 
tissue  begins  to  form  in  the  axial  area  of  the  parent  corallite,  the  daughter  corallites 
become  phaceloid. 

Increase  is  often  seen  in  hand  specimen  but  it  is  difficult  to  assess  quantitatively 
the  frequency  with  which  it  occurs.  On  one  specimen,  OUM  D508/4,  three  newly 
formed  corallites  were  seen  to  bud  themselves.  The  increments  of  vertical  growth 
between  formation  and  budding  in  these  cases  measured  1-3  cm.,  1-4  cm.  and  1-7  cm. 
Also  the  corallite  diameters  at  which  budding  may  occur  are  usually  indeterminate 
as  it  is  very  rare  to  obtain  a  cross-section  at  the  inception  of  blastogeny  when  it  is 
still  possible  to  measure  the  diameter  of  the  parent  corallite. 

A  statistical  analysis  has  been  made  of  a  total  sample  comprising  six  colonies  from 
Dyer's  quarry.  Three  of  these  colonies  have  also  been  analysed  individually.  The 
statistics  are  listed  in  Table  13  and  illustrated  graphically  in  Text-fig.  20. 

Discussion.  Specimen  UPG327  was  first  described,  with  some  other  material, 
as  Macgeea  (Thamnophyllum)  caespitosa  var.  minus  (F.  A.  Roemer)  by  Schouppe 
(1949  :  152).  Both  Fliigel  (19656  (August)  :  361)  and  Schouppe  (1956  (September)  : 
153,  footnote)  noted  that  the  Torquay  specimen  was  not  in  fact  conspecific  with 
Roemer 's  species  and  Fliigel  suggested  that,  as  a  species  in  its  own  right,  it  should 
be  called  Macgeea  (Thamnophyllum)  minima  Schouppe.  Under  article  49  (Article 
70b  does  not  apply  here)  of  the  I. C.Z.N.  (1964),  however,  Fliigel's  use  of  Roemer's 
specific  name  for  Schouppe's  misidentified  material  is  invalid  and  a  new  name  is 
needed.  With  the  permission  of  Professors  A.  von  Schouppe  and  H.  Fliigel,  "  Macgeea 
(Thamnophyllum)  minima  Schouppe  "  is  hereby  renamed,  as  a  subspecies  of  Thamno- 
phyllum germanicum,  T.  germanicum  schouppei  nom.  nov. 

In  his  original  description,  Schouppe  (1949  :  155)  stated  that  '  The  budding  is 
lateral  and  not  parricidal  (=  axial)  '  in  Macgeea  (Thamnophyllum)  caespitosum  var. 
minus.  This  is  certainly  true  of  Roemer's  species  and  Schouppe's  remark  must  have 
been  based  on  correctly  identified  material.  The  holotype,  in  fact,  gives  no  clear 
indication  of  its  style  of  increase,  although  the  evidence  from  topotypic  material  of 
the  present  subspecies  shows  that  it  must  be  axial. 

Thamnophyllum  germanicum  schouppei  is  closest  in  general  characteristics  to  the 
other  subspecies  of  T.  germanicum  described  from  the  Middle  Devonian — T.  ger- 

GEOL.  15,  5.  26§ 


264  COLONIAL  PHILLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

manicum  germanicum  nom.   nov.,   T.  germanicum  skalense   (Rozkowska)   and  T. 
germanicum  pajchelae   (Rozkowska)    (see   Rozkowska   i960  :  53).     As   Rozkowska 

(1956)  gave  data  on  corallite  diameter  and  number  of  major  septa  for  the  latter  three 
subspecies,  it  is  possible  to  compare  them  statistically  with  the  present  material 

Table  13. — Statistical  data  for  some  characters  of  Thamnophyllum  germanicum  schouppei. 

Total  sample        Colony  1  Colony  2  Colony  3 

N  153  23  50  46 

O.R.  2-9-5-7  2-9-5-2  3-Q-5-2  3-5-5-7 

X  4-19  394  4°°  4-42 

d                        s  0-59  0-51  056  059 

C.V.  14-08  12-82  13-98  !3'35 

S.E.m  0-048  o-ii  0079  0-082 

O.R.  1-7-3-7  1-7-3-3  1-8-3-1  2-1-3-7 

x  2-69  2-52  2-53  2-84 

dt         s  0-40  0-37  0-36  0-39 

C.V.  14-79  M'51  14-18  13-80 

S.E.m  0-032  0-076  0-051  0-058 

O.R.  0-55-0-70  0-59-0-70  0-55-0-65  0-62-0-69 

x  0-64  0-64  063  0-65 

dt/d       s  0016  0-023  0-020  0-021 

C.V.  2-43  3-63  3  09  3-30 

S.E.m  0-0013  00048  00028  0-0031 

O.R.  12-20  13-18  12-20  13-20 

x  16-52  15-91  16-14  16-84 

n                       s  1-22  1-24  1-51  1-09 

C.V.  7-41  7-77  9-35  6-44 

S.E.m  0-099  0-26  0-21  0-16 

O.R.  3-08-5-16  3-41-5-00  3-46-5-16  3-08-4-77 

X  3-98  4-07  4-06  3-85 

n/d                   s  0-30  0-29  0-31  0-34 

C.V.  7-51  7-00  7-72  8-78 

S.E.m  0-024  0-059  0-044  0-050 

Graphs : — 

r  0-99  0-96  0-98  0-98 

dt/d                  a  o-68  0-73  0-64  o-66 

b  —0-14  —0-34  —0-036  —0-094 

r  0-90  0-85  0-85  0-78 

n/d                   a  2- 08  2-45  2-70  1-84 

b  7-83  6-26  5-34  8-71 


n/d 


r 


—  0-94  — 0-84  —0-78  — 0-90 

a  —0-51  —0-56  —0-56  —  0-57 

b  6-n  6-29  6-30  6-38 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  265 

(Table  14  and  Text-fig.  20).  Only  T.  germanicum  pajchelae  proved  not  to  be  signifi- 
cantly different  from  T.  germanicum  schouppei  in  the  frequency  distribution  of  their 
corallite  diameters.  When  septal  number  was  plotted  against  corallite  diameter, 
however,  the  lines  of  all  three  subspecies  were  found  to  be  significantly  different  from 
that  for  T.  germanicum  schouppei. 

Qualitatively,  the  latter  differs  from  the  other  three  subspecies  by  the  absence  of 
caenogenetic  tissue  between  newly  formed  buds.  T.  germanicum  skalense  also 
differs  through  the  excessive  thickening  of  its  septa  in  the  zone  of  horseshoe  dissepi- 
ments and  T.  germanicum  pajchelae  is  remarkable  for  its  regular  and  widely  spaced 
complete  tabulae  and  general  simplification  of  its  internal  structure. 

T.  germanicum  schouppei  differs  markedly  from  the  Frasnian  subspecies  of  Thamno- 
phyllum  germanicum,  T.  germanicum  kozlowskii  (Rozkowska)  and  T.  germanicum 
superius  (Rozkowska)  (see  Rozkowska  i960  :  53),  through  the  larger  dimensions  of 
the  latter  pair.  The  mean  corallite  diameters  of  the  Frasnian  subspecies  (Rozkowska 
x957  :  93  as  T.  kozlowskii  and  T.  kozlowskii  superius)  are  10-03  and  9'2&  mm.  respec- 
tively. In  addition,  T.  germanicum  kozlowskii  is  characterized  by  a  more  complex 
dissepimentarium  than  typical  thamnophyllids,  with  one  or  two  series  of  horseshoe 
dissepiments,  and  the  tabularium  has  highly  developed  periaxial  elements.  In  the 
writer's  opinion,  this  form  deserves  full  specific  rank.     T.  germanicum  superius, 

Table  14. — Statistical  discrimination  between  some  characters  of  Thamnophyllum  germanicum. 
subspecies  and  T.  germanicum  schouppei.     Significant  values  in  bold. 

schouppei       pajchelae         skalense      germanicum 

N  153  73  69  107 


O.R. 

3-6 

2-7 

2-8 

3-10 

X 

4-14 

4-n 

5-36 

7-00 

d                      s 

0-62 

1  T3 

1-52 

i-39 

C.V. 

15-01 

27-37 

28-42 

20-81 

S.E.m 

0-050 

0-13 

0-18 

0-13 

"t"  test  against  schouppei : — 

t 

0-29 

8-46 

22-45 

O.R. 

12-20 

12-25 

10-27 

16-29 

X 

16-53 

16-90 

20-01 

23-25 

n                      s 

1  -06 

2-8l 

3-21 

2-34 

C.V. 

6-43 

16-60 

16-04 

10-07 

S.E.m 

0-086 

o-33 

o-39 

0-23 

Graph : — 

r 

0-97 

0-99 

0-98 

1  -oo 

n/d                  a 

1-71 

2-49 

2-II 

1-69 

b 

9-45 

6-65 

8-72 

n-43 

"z"  test  against  schouppei : — 

z  (slope) 

13  07 

6  00 

0-58 

z  (position) 

19  07 

266  COLONIAL  PHILLIPS  A  STRAEIDAE  FROM  S.E.  DEVON 

described  originally  as  T.  monozonatum  Soshkina  by  Rozkowska  (1953  :  14),  differs 
from  T.  germanicum  schouppei  by  the  possession  of  complete,  concave  tabulae  more 
reminiscent  of  T.  hornesi  Penecke  (as  noted  by  Rozkowska)  than  the  T.  germanicum 

§rouP- 

T.  germanicum  schouppei  is  closely  similar  in  most  respects  to  T.  caespitosum 

(Goldfuss).     The  latter,  however,  has  exclusively  lateral  increase  in  contrast  to  the 

axial  increase  of  the  former. 

Thamnophyllum  caespitosum  (Goldfuss)  sensu  lato 

Plate  15,  figs.  2-4;  Plate  16,  figs.  1,  2 

1826    Lithodendron  caespitosum  Goldfuss  :  44,  pi.  13,  fig.  4. 

1848     Cladocora  antiqua  Bronn  :  303. 

1851     Lithostrotion  antiquum  (Bronn)  Edwards  &  Haime  :  439. 

1879     Cyathophyllum  caespitosum  Goldfuss;    Quenstedt  :  509  pars. 

1881     Fascicularia  caespitosa  (Goldfuss)  Schluter  :  103,  pi.  9,  figs.  6,  7. 

1885     Cyathophyllum  caespitosum  Goldfuss ;   Freeh:    33  pars. 

1909     Phacellophyllum  caespitosum  (Goldfuss)  Gurich  :  102  pars,  pi.  31,  fig.  5. 

1935  Disphyllum  {Phacellophyllum}  caespitosum  (Goldfuss)  Lang  &  Smith  :  573,  pars,  text- 
figs.     28,  29,  pi.  35,  figs.  1,  2. 

1949  Macgeea  (Thamnophyllum)  caespitosa  (Goldfuss)  Schouppe  :  138,  pi.  9,  fig.  3,  pi.  n, 
figs.  40-43. 

1949     Phacelophyllum  caespitosum  (Goldfuss);   Stumm  :  36,  pi.  17,  figs.  11-13. 

1951  Disphyllum  {Phacellophyllum}  caespitosum  (Goldfuss);  Taylor  :  186,  pi.  3,  figs.  3a,  b, 
non  figs.  4a,  b. 

1956  Thamnophyllum  caespitosum  (Goldfuss);    Rozkowska  :  308,  text-figs.  30-32. 
1956a  Macgeea  (Macgeea)  caespitosa  (Goldfuss)  Flugel  :  54. 

1956&  Macgeea  (Macgeea)  caespitosa  (Goldfuss);   Flugel  :  361. 

1957  Thamnophyllum  caespitosum  (Goldfuss) ;    Rozkowska  :  89,  text-fig.  8. 

1958  Macgeea  ^Thamnophyllum)  caespitosum  (Goldfuss);    Schoupp6  :  227,  text-figs.  7-9. 

1963  Macgeea  (Macgeea)  caespitosa  (Goldfuss);    Schouppe  &  Stacul  :  268,  text-figs.  4,  17. 

1964  Thamnophyllum  caespitosum  (Goldfuss);   Webby  :  9,  text-figs.  ^a-d. 

But  not : 

1826     Cyathophyllum  caespitosum  Goldfuss  :  60,  pi.  19,  figs.  2a— d. 

1956    Phacellophyllum  caespitosum  (Goldfuss);   Hill  :  282,  text-fig.  192  (6). 

Diagnosis.  Phaceloid  Thamnophyllum.  Mean  tabularium  diameter  6-17  mm. 
with  18  to  22  major  septa  (Wolborough  Quarry  sample).  Major  and  minor  septa 
slightly  dilated  in  zone  of  horseshoe  dissepiments.  Dissepimentarium  regularly 
developed  with  single  outer  series  of  horizontal  plates  and  single  inner  series  of 
horseshoe  dissepiments  with  slightly  thickened  sides.  Tabularium  structure  highly 
variable  from  closely  spaced  flat-topped  domes  with  inosculating  periaxial  tabulae 
to  complete,  sagging  plates,  well  spaced  and  with  rare  subsidiary  plates.  Increase 
lateral. 

Lectotype  (selected  by  Lang  &  Smith  1935  :  573).  Original  of  Goldfuss  1826, 
pi.  13,  fig.  4  (Goldfuss  Collection,  Geologisch-Palaontologisches  Institut,  Bonn). 
Givetian ;  Bensberg,  near  Koln,  Germany. 

Material.  BM(NH)  R46167-75;  middle  Givetian  limestones,  Wolborough 
Quarry,  Newton  Abbot,  south  Devon. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  267 

Distribution.  England:  middle  and  upper  Givetian  of  the  Ilfracombe  Beds, 
north  Devon  and  west  Somerset;  Givetian  limestones,  Plymouth;  middle  Givetian 
limestones  of  Wolborough  Quarry,  Newton  Abbot,  south  Devon.  Also  common  in 
the  Givetian  of  Europe. 

Description.  Colonies  with  well  spaced,  cylindrical  corallites.  External  features 
and  colony  shape  unknown  as  the  material  is  fragmented  and  embedded  in  a  hard, 
limestone  matrix. 

In  cross-section,  the  corallites  are  circular  to  sub-circular.  The  epitheca,  which  is 
rarely  preserved  in  the  matrix,  is  about  0-05  mm.  thick.  Major  and  minor  septa  are 
slightly  and  evenly  dilated  to  o-i  mm.  thickness  for  a  length  of  1-1-25  mm-  across 
the  zone  of  horseshoe  dissepiments.  The  minor  septa  appear  to  penetrate  slightly 
into  the  tabularium  whilst  the  major  septa,  thin  and  often  slightly  sinuous,  continue 
towards  the  axis.     The  latter  are  a  half  to  two-thirds  the  corallite  radius  in  length. 

The  traces  of  the  horseshoe  dissepiments  form  two  strong  internal  walls,  convex 
towards  each  other,  in  cross-section. 

In  longitudinal-section,  the  dissepimentarium  comprises  an  outer  series  of  hori- 
zontal plates  and  an  inner  series  of  horseshoe  dissepiments,  their  respective  widths 
in  the  ratio  1  :  1-3.  The  plates  of  the  outer  series  may  be  flat  or  very  slightly  curved 
and  are  exclusively  uniserial.  Their  spacing  varies  between  0-25  mm.  and  0-65  mm. 
The  horseshoe  dissepiments  are  also  exclusively  uniserial  and  fairly  uniform  in  size 
and  shape.  They  average  0-3  mm.  high.  The  crest  of  each  horseshoe  is  thin  but 
the  sides  are  moderately  and  evenly  thickened  up  to  o-oi  mm. 

The  axial  tabulae  are  complete  or  incomplete,  usually  closely  spaced  and  arched 
with  a  wide  flat  crest.  Periaxial  elements  are  usually  well  developed  as  small 
arched  plates  inosculating  with  the  main  tabulae.  Tabularium  structure  is,  however, 
very  variable  and  includes  corallites  with  wide  flat  tabulae  with  scattered  periaxial 
plates  and  rarely,  well  spaced,  dominantly  complete,  saucer-shaped  tabulae.  The 
vertical  spacing  of  plates  in  the  axis  of  the  tabularium  varies  from  14  per  cm.  when 
the  structure  is  complex,  to  10  per  cm.  when  the  structure  is  simple. 

Increase,  only  observed  in  a  few  cases,  is  lateral  with  a  high  angle  of  divergence 
between  parent  and  bud. 

It  was  possible  to  measure  only  a  few  corallites  accurately  so  that  no  attempt  has 
been  made  to  analyse  the  figures  in  detail.     Data  obtained  are  given  in  Table  15&. 

Discussion.     See  under  Thamnophyllum  caespitosum  paucitabulatum  (p.  268). 


Thamnophyllum  caespitosum  paucitabulatum  subsp.  nov. 

Plate  15,  fig.  1;  Plate  16,  figs.  3-7;  Plate  17,  figs.  1-3 

Derivation  of  name.  The  name,  pauci-  (paucus  L.  =  few)  tabulatum,  refers  to 
the  simple,  relatively  wide  spaced  tabulae  characteristic  of  the  subspecies. 

Diagnosis.  Thamnophyllum  caespitosum  with  tabularium  composed  of  flat, 
slightly  arched,  or  saucer-shaped  tabulae  with  rare  periaxial  elements.  Mean 
tabularium  diameter  5-68  mm.  with  16  to  21  major  septa  (Topotype  sample). 


268  COLONIAL  PH  I  LLI  PS  ASTR  AEID  AE  FROM  S.E.  DEVON 

Holotype.  BM(NH)  R46165;  upper  Givetian  limestones;  Lummaton  Quarry, 
Torquay,  south  Devon. 

Material.     BM(NH)  R46160-6. 

Distribution.     Type  locality  only. 

Description.  Subphaceloid  colonies  of  indeterminate  size.  Corallites  cylindri- 
cal, subparallel  and  fairly  well  spaced.  External  features  are  unknown  as  the  material 
is  embedded  in  a  hard  limestone  matrix. 

The  structural  details  are  as  for  Thamnophyllum  caespitosum  above  with  the 
following  exceptions  and  additions. 

In  longitudinal-section,  the  tabulae  are  usually  complete.  They  may  be  slightly 
arched  with  wide  flat  crests,  when  small  arched  periaxial  elements  are  poorly 
developed.  More  commonly,  the  tabulae  are  flat  or  saucer-shaped  with  very  rare 
flat  or  slightly  arched  subsidiary  plates.  There  are  between  8  and  12  tabulae  per 
cm.;  spacing  of  the  plates  increases  from  corallites  with  arched  tabulae  to  those 
with  saucer-shaped  tabulae. 

Increase  is  exclusively  lateral,  occasionally  with  two  offsets  developed  at  the  same 
level.  A  small  amount  of  extradissepimental  tissue  is  always  developed  between 
the  parent  and  the  bud  but  the  angle  of  divergence  is  usually  high,  restricting  the 
plocoid  stage  to  a  minimum.  Measured  corallites  showing  the  early  stages  of  increase 
are  6-5  and  6-6  mm.  in  diameter. 

A  statistical  analysis  has  been  made  of  the  sample  from  Lummaton  Quarry. 
Individual  colonies  cannot  unfortunately  be  distinguished.  The  statistics  are  listed 
in  Table  15a  and  illustrated  graphically  in  Text-fig.  21a. 

Discussion.  Previous  descriptions  of  Thamnophyllum  caespitosum  (Lang  & 
Smith  1935  :  573;  Rozkowska  1956  :  308;  Webby  1964  :  9)  all  mention  tabularia 
composed  of  flat  or  slightly  arched  tabulae  with  well  developed  periaxial  elements. 
No  great  variation  in  this  structure  is  mentioned.  In  fact  Flugel  (19566  :  361)  and 
Schouppe  &  Stacul  (1963  :  268)  have  placed  this  species  in  Macgeea  on  the  basis  of 
the  complex  tabularium  structure  (see  discussion  of  Thamnophyllum,  p.  259).  The 
material  collected  from  Wolborough  and  Lummaton  Quarries  in  south  Devon, 
however,  shows  considerable  variation  in  the  shape  and  distribution  of  the  tabulae. 
The  sample  from  Wolborough  has  a  particularly  wide  range  although  the  majority 
of  the  corallites  display  the  tabularium  structure  considered  typical  of  T.  caespitosum 
sensu  stricto.  At  Lummaton,  on  the  other  hand,  the  typical  form  is  not  represented 
and  in  most  cases  the  tabularia  have  wide  spaced  flat  or  saucer-shaped  tabulae  with 
only  rare  periaxial  elements.  Although  in  other  respects  the  two  samples  are 
virtually  identical,  the  range  in  tabularium  structure  is  so  striking  and,  in  view  of 
the  taxonomic  weight  hitherto  placed  on  this  character,  so  important  that  the  erec- 
tion of  a  new  subspecies,  T.  caespitosum  paucitabulatum,  for  the  Lummaton  material 
is  felt  to  be  justified. 

T.  caespitosum  caespitosum  is  interpreted  strictly  in  terms  of  the  lectotype  (see 
Lang  &  Smith  1935  :  573)  and  includes  the  material  described  by  Rozkowska  and 
Webby.     The  highly  variable  middle  Givetian  Wolborough  material,  however,  is 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

10 


269 


f     10 


395  475  5  55  6  35 

Q      (0  4  mm  class    intervals) 


50 


dt 


40 


30 


(a) 


715 


365  445  525  6  05  6  85 

d     (0-4  mm  class   intervals) 


5  0  r 


dt 

(mm) 


40 


30 


2  0  *-•->- 

22 


21 
20 
19 
18 


17 
50 


45 





%AO 


35   • 


30 


40 


50 


60 


(b) 


Fig.    21.     a,    Thamnophyllum    caespitosum    pancitabulatum.     b,  Peneckiella    salternensis. 


270 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 


Table  15. — a.  Statistical  data  for  some  characters  of  Thamnophyllum  caespitosum  paucitabulatum. 

b.  Statistical  data  for  some  characters  of  Thamnophyllum  caespitosum  (sensu  lato) .     c.  Statistical 

data  for  some  characters  of  Peneckiella  salternensis. 


(a) 

(b) 

(c) 

Total  sample 

Total : 

sample 

Total  sample 

N 

62 

12 

33 

O.R. 

4-5-6-9 

5-5- 

-7-2 

4-0-6-3 

X 

5-68 

6 

•17 

5-28 

d 

s 

o-55 

°53 

G.V. 

9-66 

IO-II 

S.E.m 

0-070 

0-093 

O.R. 

3-Q-4-9 

3-5" 

-4-8 

2-7-4-3 

X 

3-70 

4 

•14 

3-5i 

dt 

s 

0-40 

0-38 

C.V. 

io-88 

10-87 

S.E.m 

0-051 

0-066 

O.R. 

0-62-0-71 

0 • 63-0 • 68 

X 

0-65 

0-67 

dt/d 

s 

0-021 

O-II 

C.V. 

3-27 

i-68 

S.E.m 

0-0027 

0 • 0020 

O.R. 

16-21 

18- 

-22 

17-21 

X 

I8-92 

20 

00 

19-91 

n 

s 

o-75 

0-84 

C.V. 

3-95 

4-21 

S.E.m 

0-095 

0-15 

O.R. 

2-61-4-26 

4-28-3-33 

X 

3-35 

3 

26 

3-80 

n/d 

s 

0-29 

0-28 

C.V. 

8-78 

7"34 

S.E.m 

0-037 

0-049 

Graphs : — 

r 

o-95 

0-99 

dt/d 

a 

o-73 

0-72 

b 

-0-47 

—  0-27 

r 

0-48 

o-8r 

n/d 

a 

1-36 

i-57 

b 

11-19 

"•53 

n/d/ 
/d 

r 

—  0-90 

-0-94 

a 

-o-54 

—0-52 

/ 

b 

6-40 

6-55 

COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  271 

assigned  to  T.  caespitosum  sensu  lato,  from  which  it  is  thought  the  upper  Givetian  T. 
caespitosum  pancitabulatum  from  Lummaton  directly  descended. 

Quantitatively,  T.  caespitosum  paucitabulatum  has  a  slightly  lower  mean  corallite 
diameter  than  that  for  T.  caespitosum  sensu  lato  although  the  latter  has  insufficient 
measurements  for  accurate  comparison  (Tables  15a,  b).  Figures  available  for  T. 
caespitosum  caespitosum — '  close  to  6  mm.  '  corallite  diameter  with  '  usually  18 
to  19  '  major  septa  (Webby  1963  :  10)  and  '  corallites  average  about  6  mm.  in 
diameter  and  have  about  20  major  septa  '  (Rozkowska  1956  :  308) — compare  closely 
with  those  for  the  Wolborough  material.  Lang  &  Smith  (1953  :  574),  however, 
record  fewer  major  septa — 16  to  18 — with  an  average  corallite  diameter  of  6  cm. 
(sic)  in  the  lectotype  of  T.  caespitosum  caespitosum. 

T.  caespitosum  is  structurally  most  similar  to  the  T.  germanicum  group  but  is 
distinguished  from  it,  and  all  other  species  of  Thamnophyllum,  by  the  possession  of 
lateral  budding. 

Thamnophyllum  spp. 

Material.  BM(NH)  R46181,  middle  Givetian  limestones  (see  Middleton  1959), 
Shinner's  Bridge  Quarry,  near  Dartington  (SX  78906225);  BM(NH)  R46178, 
Givetian  limestones,  road  cutting  immediately  south  of  junction  of  Babbacombe 
Road  with  Acre  Lane,  Torquay  (SX  93186477) ;  BM(NH)  R46177,  Givetian  lime- 
stones, disused  quarry  on  Teignmouth  Road,  Torquay  (SX  91126553) ;  BM(NH) 
R46179-80,  Middle  Devonian  (PGivetian)  thin  bedded  limestones,  80  ft.  above  sea 
level  in  cliff  at  northern  end  of  Redgate  Beach,  Torquay  (SX  93516494).  All  south 
Devon. 

Discussion.  These  fragments,  which  do  not  all  belong  to  the  same  species  or 
subspecies,  can  nevertheless  be  placed  in  either  Thamnophyllum  caespitosum  or  T. 
germanicum.  Unfortunately  none  of  them  gives  evidence  of  the  mode  of  increase 
which  is  critical  in  distinguishing  between  the  two  species.  Further  material  may 
eventually  enable  the  accurate  determination  of  these  specimens. 

Genus  PENECKIELLA  Soshkina  1939 

partim  1939     Peneckiella  Soshkina  :  23. 

partim  1939     Disphyllum;  Hill  :  224. 

partim  1949    Peneckiella;  Soshkina  :  141.  1 

partim  1949     Macgeea  (Thamnophyllum);  Schouppe  :  115. 

partim  1950    Phacellophyllum  (Phacellophyllum) ;  Wang  :  219. 

Soshkina  :  103. 

Soshkina  :  103. 

Soshkina  :  32. 

Hill  :  25. 

Schoupp6  :  153. 

Fliigel  :  55. 

Fliigel  :  355. 

Hill  :  282. 

Schouppd  :  229. 


partim  1951     Peneckiella 

partim  1952     Peneckiella 

partim  1954    Peneckiella 

partim  1954a  Peneckiella 

1956    Peneckiella 

1956a  Peneckiella 

1956b  Peneckiella 

partim  1956    Peneckiella 

1958    Peneckiella 


1959     Acinophyllum  McLaren  :  22. 


272  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

partim  1959  Phacellophyllum;  McLaren  :  28. 

i960  Peneckiella;  Rozkowska  :  29. 

i960  Sudetia  Rozkowska  :  35. 

partim  1965  Peneckiella;  Strusz  :  555. 

Diagnosis.  Phaceloid  or  dendroid  rugose  corals.  Septa,  major  and  minor,  may 
be  lightly  carinate  and  dilated  in  dissepimentarium.  Characterized  by  one  or  two, 
seldom  more,  series  of  dissepiments  of  variable  form  but  always  including  peneckiel- 
loid  dissepiments.  Horseshoe,  flat  and  sigmoidal  dissepiments  may  also  be  present. 
Tabulae  complete  or  incomplete,  frequently  with  flat-topped  domes.  Increase 
usually  lateral  but  may  be  axial. 

Type  species.  Diphyphyllum  minus  F.  A.  Roemer  (1855  :  29,  pi.  6,  figs,  iia-c). 
Frasnian,  Ibergerkalk;   Winterberg  near  Bad  Grund,  Harz,  Germany. 

Distribution.  Uppermost  Givetian  and  Frasnian  of  Europe;  Pupper  Lower 
Devonian  to  Frasnian  of  Australia;  ? Lower,  Middle  and  Upper  Devonian  of  North 
America. 

Discussion.  In  her  original  diagnosis,  Soshkina  (1939  :  23)  included  '  simple  or 
composite,  fasciculate  and  massive  '  corals  and  described  the  dissepimentarium  as 
composed  of  '  one  row  of  regularly  spaced  vesicles  .  .  .  sometimes  flattened  from 
above  '.  Soshkina  (1954  :  32)  later  gave  a  slightly  different  diagnosis  for  the  genus 
which  Flugel  (19566  :  355)  mistranslated  as  excluding  massive  forms  from  Peneckidla. 
Schouppe  (1958  :  191)  pointed  out  Fliigel's  mistake  but  himself  concluded  that 
Peneckiella  should  be  correctly  defined  (1958  :  192)  as  excluding  massive  forms. 
Both  Flugel  (19566)  and  Schouppe  (1958)  gave  extended  discussions  of  this  genus 
and  concluded  that  it  is  characterised  basically  by  a  single,  rarely  double  row  of 
horseshoe  dissepiments  only  in  the  dissepimentarium. 

It  is  clear,  however,  from  the  holotype  of  Peneckiella  minor,  type  species  of 
Peneckiella,  that  it  is  peneckielloid  (see  Rozkowska  i960  :  32)  and  not  true  horseshoe 
dissepiments  that  are  characteristic  of  the  genus.  No  true  horseshoe  dissepiments  can 
be  positively  identified  in  the  holotype  (Plate  17,  figs.  4,  5),  although  they  do  occur 
in  a  subsidiary  role  in  the  dissepimentaria  of  some  topotype  specimens. 

McLaren  (1959  :  22)  in  the  discussion  of  his  new  genus  Acinophyllum,  remarked 
that,  from  Freeh's  (1885  :  34,  pi.  1,  figs.  3,  3a,  36)  account  of  Cyathophyllum  minus 
(F.  A.  Roemer),  which  was  based  on  Roemer's  specimen,  '  it  would  appear  likely 
that  Acinophyllum  simcoense  is  congeneric  with  Diphyphyllum  minus  and  therefore 
with  Peneckiella.  But  D.  minus  has  been  described  several  times  since  Freeh  and 
on  no  occasion  has  the  description  agreed  closely  with  his.  '  In  fact  Freeh's  drawing 
(pi.  1,  fig.  3a)  more  accurately  portrayed  the  dissepimental  structure  of  D.  minus 
(except  for  the  divergence  of  the  septal  trabeculae)  than  either  Fliigel's  (19566, 
text-fig.  ic)  or  Schouppe's  (1958,  text-figs.  13,  14)  illustrations. 

Dr.  W.  A.  Oliver  Jr.  has  kindly  sent  the  writer  two  small  fragments  (BM(NH) 
R463661/-2)  from  a  large  colony  of  Acinophyllum  simcoense  from  the  Bois  Blanc 
Formation  (south  quarry,  Haldimand  Quarries  and  Construction  Ltd.,  north  east 
of  Hagarsville,  Ontario) .  Slides  cut  from  this  material  show  well  developed  peneckiel- 
loid dissepiments  in  longitudinal-section  extremely  similar  to  those  of  P.  minor. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  273 

Where  subsidiary  dissepiments  are  developed,  however,  these  are  normal  globose 
vesicles;  no  horseshoe  dissepiments  have  been  seen.  The  septal  carination  in 
cross-section  is  very  variable  and  smooth  septa  do  occur.  Certainly  the  presence  or 
absence  of  carinae  cannot  be  considered  to  be  of  great  significance  here  at  the  generic 
level.  On  all  other  characters,  the  specimen  of  A.  simcoense  can  be  placed  with 
little  doubt  in  Peneckiella  and  Acinophyllum  is  thus  considered  a  junior  synonym 
of  that  genus. 

Rozkowska  (1960  :  35)  erected  a  new  genus  and  species,  Sudetia  lateseptata  which 
she  considered  (i960  :  50)  to  be  a  direct  phylogenetic  descendant  of,  and  closely 
related  to  Peneckiella.  The  author  is  of  the  same  opinion  but  believes  that  Sudetia 
does  not  warrant  separate  generic  status. 

Peneckiella,  as  suggested  by  Rozkowska  (i960),  apparently  evolved  from  Thamno- 
phyllum,  chiefly  by  the  fusion  and  modification  of  the  horseshoe  and  flat  dissepiments 
characteristic  of  the  latter,  leading  to  the  diversification  of  dissepimental  types  found 
in  Peneckiella.  Besides  the  characteristic  peneckielloid  dissepiments,  forms  such  as 
P.  minor  kunthi  (Dames)  (see  Rozkowska  i960,  text-fig.  27),  P.  mesa  (Hill)  (see 
Strusz  1965,  text-fig.  156)  and  P.  salternensis  sp.  nov.  (see  p.  274),  include  varying 
proportions  of  horseshoe,  flat  and  sigmoidal  dissepiments  in  their  dissepimentaria. 
In  fact,  a  morphological  series  can  apparently  be  traced  from  a  horseshoe  and  flat 
dissepimental  pair,  through  sigmoidal  to  peneckielloid  dissepiments.  P.  minor 
sensu  stricto,  with  a  fairly  uniform  peneckielloid  dissepimentarium  and  only  rare 
horseshoe  dissepiments  would  appear  to  be  an  advanced  form  and  P.  lateseptata, 
as  suggested  by  Rozkowska,  a  late  stage  form  in  this  evolutionary  trend. 

Peneckiella  salternensis  sp.  nov. 

Plate  18,  figs.  1-4 

1965     Peneckiella  cf.  minor  (Roemer);  Scrutton:    188,  text-fig.  1. 

Derivation  of  name.     After  the  type  locality  in  Saltern  Cove. 

Diagnosis.  Phaceloid  Peneckiella.  Mean  tabularium  diameter  3-51  mm.,  mean 
corallite  diameter  5-28  mm.  with  17  to  21  major  septa  (topotype  sample).  Septa 
slightly  and  variably  dilated;  may  be  weakly  carinate.  Dissepimentarium  domi- 
nated by  peneckielloid  dissepiments,  but  with  horseshoe,  flat  and  sigmoidal  dissepi- 
ments also  present.  Tabulae  regularly  developed  in  form  of  wide,  flat-topped  domes. 
Increase  lateral. 

Holotype.  OUM  D548.  Frasnian;  thick  bedded  limestone  immediately  above 
igneous  rock;  southern  end  of  Saltern  Cove  (SX  89505842),  near  Paignton,  south 
Devon. 

Material.  Saltern  Cove  (main  Peneckiella  horizon) ;  measured  material:  OUM 
D501,  OUM  D546-48,  OUM  D550-51,  OUM  D553;  additional  material:  OUM 
D549,  OUM  D552.     Saltern  Cove  (thin  bedded  limestones) :   OUM  D554. 

Distribution.  Frasnian;  type  horizon  (main  Peneckiella  horizon)  and  from 
thin  bedded  limestones  in  the  sequence  immediately  above,  southern  end  of  Saltern 
Cove,  south  Devon. 


274  COLONIAL  PH  ILLIPS  ASTR  AEI  D  AE  FROM  S.E.  DEVON 

Description.  Colonies  are  phaceloid  with  circular  to  sub-circular,  close  spaced 
corallites.  External  features  are  unknown  as  the  specimens  are  preserved  in  a 
tough  matrix.     The  epitheca  is  o-r  mm.  or  slightly  less  in  thickness. 

The  major  and  minor  septa  are  variably  dilated  in  the  dissepimentarium,  occasion- 
ally strongly  spindle-shaped  but  showing  all  gradations  to  a  virtually  unthickened 
state.  Their  thickness  ranges  from  about  0-3  mm.  to  o-i  mm.  or  slightly  less.  The 
septa  may  be  straight  or  zigzagged  in  this  zone,  and  are  sometimes  variably  carinate. 
Carinae  are  yardarm  on  straight  septa  and  xyloid  on  zigzagged  septa. 

Minor  septa  end  at  the  tabularium  junction  but  the  major  septa,  slightly  attenuated, 
continue  into  the  tabularium  with  a  thickness  of  about  0-05  mm.  or  less.  The  major 
septa  may  be  straight  to  slightly  sinuous  in  the  tabularium  and  normally  extend 
about  half  way  to  the  axis.  Septal  length  is  somewhat  variable,  however,  and 
rarely  they  may  more  or  less  reach  the  axis  or,  conversely,  only  just  penetrate  into 
the  tabularium.     The  axial  ends  of  the  major  septa  may  rarely  be  slightly  thickened. 

The  traces  of  one  or  two  rows  of  dissepiments,  uniserial  between  adjacent  septa, 
may  be  seen  in  cross-section.  The  tabularium  junction  can  be  easily  distinguished 
but  it  is  not  usually  strongly  defined. 

In  longitudinal-section,  the  dissepimentarium  is  composed  of  one  or  two  series  of 
highly  variable  dissepiments.  The  peneckielloid  form  dominates  when  the  dissepi- 
mentarium is  usually  but  not  always  uniserial.  The  occurrence  of  horseshoe  dissepi- 
ments may  be  accompanied  by  peripheral  flat  dissepiments  but  normally  the  latter 
are  very  rare;  horseshoes  always  occur  in  the  inner  row  of  biserial  parts  of  the 
dissepimentarium.  Sigmoidal  dissepiments  occur  with  about  the  same  frequency  as 
horseshoe  dissepiments.  The  ratio  of  peneckielloid  dissepiments  to  horseshoe 
dissepiments  is  highly  variable  from  corallite  to  corallite,  ranging  between  extremes 
of  2  :  1  and  10  :  1.  The  various  dissepimental  types  are  randomly  distributed  up 
the  length  of  the  corallites  and  the  vertical  spacing  of  the  dissepiments  may  vary 
from  o-i  to  0-5  mm. 

Tabularium  structure  is  very  constant,  consisting  of  regularly  and  closely  spaced, 
wide,  flat-topped  domes.  Periaxial  tabulae  in  the  form  of  steeply  dipping  peripheral 
vesicles  may  sometimes  occur.     There  are  about  40  tabulae  in  1  cm. 

Increase  is  lateral  corresponding  closely  to  the  '  thamnophylloid  lateral '  type 
described  by  Rozkowska  (i960  :  31).  The  diameters  of  parent  corallites  displaying 
increase  range  from  5-3  to  5-6  mm. 

A  statistical  analysis  of  the  Saltern  Cove  material  has  been  made  but  unfortunately 
individual  colonies  could  not  be  distinguished.  The  statistics  are  listed  in  Table  15c 
and  illustrated  graphically  in  Text-fig.  2xb. 

Discussion.  Peneckiella  salternensis  can  be  distinguished  from  P.  lateseptata 
(Rozkowska),  P.  nalivkini  Soshkina  and  P.  achanaiensis  Soshkina  on  dissepimental 
character  alone.  The  latter  three  are  all  characterized  by  almost  exclusively 
uniserial  dissepimentaria  of  uniformly  developed  peneckielloid  dissepiments.  P. 
minor  minor  (Roemer)  apparently  has  a  somewhat  less  regular  dissepimentarium, 
with  the  presence  of  occasional  horseshoes.  The  tabularium,  however,  although 
sometimes  developing  dome-shaped  plates,  shows  tabulae  irregularly  developed  and 
frequently  depressed  in  the  axial  area.     This  is  in  strong  contrast  to  the  regularly 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  275 

developed  dome-shaped  tabulae  characteristic  of  P.  saltemensis.  In  addition,  data 
given  by  Flugel  (19566  :  359)  for  P.  minor  minor  (Table  16)  shows  that  subspecies 
to  be  smaller,  with  fewer  septa  than  P.  saltemensis. 

P.  minor  kunthi  (Dames),  P.  mesa  (Hill)  and  P.  boreensis  Struszhave  dissepimentaria 
of  similar  complexity  to  P.  saltemensis.  P.  minor  kunthi,  however,  has  flat,  or  more 
commonly,  incomplete,  axially  depressed  tabulae  and  is  smaller  (Table  16)  than  P. 
saltemensis.  P.  mesa  and  P.  boreensis  have  flat-topped  domes  in  their  tabularia, 
but  both  Australian  species  have  axial  increase  in  contrast  to  the  lateral  increase  in 
P.  saltemensis.  In  addition,  P.  saltemensis  is  larger  than  P.  mesa  and  P.  boreensis 
(see  Table  16)  and  the  latter  is  further  distinguished  by  excessive  dilatation  of  the 
septa  in  the  dissepimentarium. 

Table  16. — Quantitative  comparison  of  some  species  and  subspecies  of  Peneckiella. 

d  dt  n 


P.  saltemensis 

O.R. 

4-0-6-3 

2-7-4-3 

17-21 

Paignton,  S.  Devon, 
England. 

X 

5-28 

3-51 

19-91 

P.  minor  minor 

O.R. 

3-5-4-9 

2-9-3-9 

15-18 

Harz,  Germany. 

1 

X 

3-9 

P.  minor  minor 

O.R. 

3-5-5-6 

2-9-4-2 

14-18 

Antitaurus,  Turkey. 

1 

X 

4-8 

P.  minor  kunthi 

O.R. 

2-5-4-8 

12-18 

Mokrzesz6w,  Poland. 

2 

X 

P.  mesa 

O.R. 

1-4-8-4 

1-0-3-9 

11-23 

Wellington,  N.S.W., 
Australia. 

3 

X 

3-6 

2-12 

17-00 

P.  boreensis 

O.R. 

0-9-10-1 

0-7-6-5 

3-28 

Molong,  N.S.W., 
Australia. 

3 

X 

4-5 

2-8 

16-00 

Data  from    1  Flugel,  1956b:  359. 

2  R6zkowska,  i960:  29 

3  Strusz,  1965:557 

,  562. 

Unfortunately  data  given  for  other  species  and  subspecies  of  Peneckiella  are  not 
in  a  form  allowing  statistical  comparison  with  the  present  material.  Nevertheless 
they  afford  a  valuable  general  indication  of  quantitative  relationships  and  are  there- 
fore listed  in  Table  16. 

REFERENCES 

Blainville,  H.  M.  D.  de     1834.     Manuel  d'Actinologie  ou  de  Zoophytologie.     viii  +  695  pp. 

Atlas.     Paris. 
Bronn,  H.  G.     1848-9     Index  Palaeontologicus.     1   (1),  lxxxiv  +  1-775   (1848);    1  (2),   777- 

1381  +  [1]  (1848);   2,  1-980  (1849).     Stuttgart. 


276  COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON 

Bulvanker,  E.   Z.     1958.     Devonskie  chetyrekhluchevye  korally  okrain  Kuznetskogo  basseyna. 

2.12,  pp.  93  pis.  VSEGEI,  Leningrad.     [In  Russian]. 
Chilingar,  G.  V.     1956.     Review:  About  dynamics  of  septal  development  in  tetracorals  during 

ontogeny,  by  K.  G.  Voynovskiy-Kriger.     J.  Paleont.,  Chicago,  30  :  406-411,  9  text-figs. 
Dana,  J.  D.     1846.     Genera  of  fossil  corals  of  the  family  Cyathophyllidae.     Am.  J.  Sci.  Arts., 

New  Haven  (2),  1  :  178-189,  text-figs.  1-5. 
Dineley,   D.   L.     1 961.     The  Devonian  system  in  South  Devonshire.     Fid.  Stud.,   London, 

1  :  1 21-140,  6  text-figs. 
Dineley,  D.  L.  &  Rhodes,  F.  H.  T.     1956.     Conodont  horizons  in  the  west  and  south-west  of 

England.     Geol.  Mag.,  London,  93  :  242-248,  text-figs.  1-5. 
Edwards,  H.  M.   &  Haime,   J.     1850.     A  Monograph  of  the  British  Fossil  Corals,   Part  1. 

Introduction,  etc.  :  i-lxxxv  +  1-71,  pis.  1-11.     Palaeontogr.  Soc.  [Monogr.],  London. 
1851.     Monographie  des  Polypiers  fossiles  des  Terrains  palaeozo'iques.    Arch.  Mus.  natn. 

Hist.  nat.  Paris,  5  :  1-501,  pis.  1-20. 

1852.     A  Monograph  of  the   British  Fossil  Corals,   Part  3.     Corals  from   the  Permian 

Formation  and  the  Mountain  Limestone  :  147-210,  pis.  31-46.     Palaeontogr .  Soc.  [Monogr.], 
London. 

1853.     A  Monograph  of  the  British  Fossil  Corals,   Part  4.     Corals  from  the  Devonian 

Formation  :  211-244,  pls-  47_5°-     Palaeontogr.  Soc.  [Monogr.],  London. 

1855.     A  Monograph  of  the  British  Fossil  Corals,  Part  5.     Corals  from  the  Silurian  Forma- 
tion :  245-299,  pis.  57-72.     Palaeontogr.  Soc.  [Monogr.],  London. 

Ehlers,  G.  M.  &  Stumm,  E.  C.  1953.    Species  of  the  Tetracoral  genus  Billingsastraea  from  the 

Middle  Devonian  of  New  York  and  other  regions.     Bull.  Buffalo  Soc.  nat.  Sci.,  21  :  1-11, 

pis.  1-6. 
Elliott,  G.  F.     1961.     A  new  British  Devonian  alga,  Palaeoporella  lummatonensis ,  and  the 

brachiopod  evidence  of  the  age  of  the  Lummaton  Shell-Bed.      Proc.  Geol.  Ass.,  London, 

72  :  251-260,  pis.  9,  10. 
Etheridge,  R.     1899.     Descriptions  of  new  or  little  known  Victorian  Palaeozoic  or  Mesozoic 

fossils,  No.  1.     Prog.  Rept.  geol.  Surv.  Vict.,  Melbourne,  11  :  30-36,  pis.  A,  B. 
Fleming,  J.     1828.     A  History  of  British  Animals,     xxiii  +  565  pp.  Edinburgh. 
Flugel,  H.     1956a,  Die  "  Sandsteinfazies  "  des  Mitteldevons  von  Graz.     Sber.  Akad.   Wiss. 

Wien,  math.-nat.  Kl.,  1956  :  45-57,  2  tables. 

19566.     Kritische    Bemerkungen    zum    Genus    Peneckiella    Soshkina.     Neues    Jb.    Geol. 

Paldont.  Mh.,  Stuttgart,  1956,  8  :  355-365,  3  text-figs. 

1958.     Korallen    und    Stromatoporen    aus    den    Gerollen    der    Karbonkonglomerate   des 

Ljubljana-Feldes.     Razpr.   Slov.   Akad.    Znan.    Umet.,    Ljubljana    (4),    4  :  623-629,    text- 
fig.  1. 

1959.     Zur  Kenntnis  der  Typen  von  Favistella  (Dendrostella)  trigemme  trigemme  (Quenst., 

1881)  und  Thamnophyllum  trigeminum  trigeminum  Penecke  1894.     Neues  Jb.  Geol.  Paldont. 
Mh.,  Stuttgart,  1959,  3  :  1 13-120. 

Fontaine,  H.     1961.     Les  Madreporaires  paleozoiques  du  Viet-Nam,  du  Laos  et  du  Cambodge. 
Archs  geol.  Viet-Nam,  Saigon,  5  :  1-276.     Atlas,  pis.  1-35. 

1964.     Madr6poraires  paleozoiques  du  Viet-Nam,  du  Cambodge,  du  Laos  et  du  Yunnan. 

Nouvelles    determinations    et    notes    bibliographiques.     Archs    geol.     Viet-Nam,    Saigon, 
6  :  75-92,  pi.  8. 

Frech,    F.     1885.     Die    Korallenfauna   des  Oberdevons   in   Deutschland.     Z.    dt.   geol.    Ges., 

Berlin,  37  :  21-130,  pis.  1-11. 
Glinski,  A.     1955.     Cerioide  Columnariidae   (Tetracoralla)   aus  dem  Eifelium  der  Eifel  und 

des  Bergischen  Landes.     Senckenberg.  leth.,  Frankfurt  a.M.,  36  :  73-114,  pis.  1,  2. 
Goldfuss,    G.   A.     1826-33.     Petrefacta  Germaniae,    1  :  1-76,   pis.    1-25    (1826);    77-164,   pis. 

26-50(1829);    165-240,  pis.  51-71  (1831);    241-252(1833).     Diisseldorf. 
Gurich,  G.     1896.     Das  Palaeozoicum  des  Polnischen  Mittelgebirges.      Verhandl.  Russ.-kais. 

Min.  Ges.  St.  Petersb.  (2)  32  :  vi  -f-  1-539,  pis.  1-15. 
1909.     Leitfossilien,  2.  Devon  :  97-199,  pis.  29-52.     Berlin. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  277 

Hill,  D.     1935-     British  terminology  for  rugose  corals.    Geol.  Mag.,  London,   72  :  481-519, 
zi  text-figs. 

1939-     The  Devonian   Rugose  Corals  of  Lilydale  and  Loyola,   Victoria.     Proc.  R.  Soc. 

Vict.,  Melbourne  (n.s.)  51,  2  :  219-256,  pis.  13-16. 

1940.     The  Lower  Middle  Devonian  Rugose  Corals  of  the  Murrumbidgee  and  Goodradig- 

bee  Rivers,  N.S.W.     /.  Proc.  R.  Soc.  N.S.W.,  Sydney,  74  :  247-276,  pis.  9-1 1. 

1942a.     The  Lower  Devonian  Rugose  Corals  from  the  Mt.  Etna  Limestone,  Qld.  Proc. 

R.  Soc.  Qd.,  Brisbane,  54  :  13-22,  pi.  1. 

19426.     The  Devonian  Rugose  Corals  of  the  Tamworth  District,  N.S.W.    /.  Proc.  R.  Soc. 

N.S.W. ,  Sydney,  76  :  142-164,  pis.  2-4. 

1942c.     Middle  Palaeozoic  Rugose  Corals  from  the  Wellington  District,  N.S.W.     /.  Proc. 

R.  Soc.  N.S.W.,  Sydney,  76  :  182-189,  pis.  5,  6. 

1954a.     Coral  Faunas  from  the  Silurian  of  New  South  Wales  and  the  Devonian  of  Western 

Australia.     Bull.  Bur.  Miner.  Resour.  Geol.  Geophys.  Aust.,  Melbourne,  23  :  1-51,  pis.  1-4. 

I954&.     Devonian  Corals  from  Waratah  Bay,  Victoria.     Proc.  R.  Soc.   Vict.,  Melbourne, 

(n.s.),  66  :  105-118,  pis.  6-9. 

1956.     Rugosa.     In  Moore,  R.  C.  (Ed.)  Treatise  on  Invertebrate  Paleontology,  F  :  233-324, 

text-figs.  165-219.     Lawrence,  Kansas. 
House,  M.   R.     1963.     Devonian  ammonoid  successions  and  facies  in  Devon  and  Cornwall. 

Q.  J I  geol.  Soc.  Lond.,  119  :  1-27,  pis.  1-4. 
House,  M.   R.   &  Selwood,   E.   B.     1965.     Palaeozoic  palaeontology  in  Devon  and  Cornwall. 

In  Hosking,  K.  F.  G.  &  Shrimpton,  G.  J.  (Eds.)  Present  views  of  some  aspects  of  the  geology 

of  Cornwall  and  Devon  :  45-86,  pis.  1-4.     Penzance. 
Imbrie,  J.     1956.     Biometrical  methods  in  the  study  of  Invertebrate  fossils.     Bull.  Am.  Mus. 

nat.  Hist.,  New  York,  108  :  211-252,  10  text-figs. 
International  Code  of  Zoological  Nomenclature.     1964.     xx  +  176  pp.     London. 
Jukes-Browne,  A.  J.     1906.     The  Devonian  Limestones  of  Lummaton  Hill,  near  Torquay. 

Proc.  Geol.  Ass.,  London,  19  :  291-302,  1  text-fig. 
Kato,  M.     1963.     Fine  skeletal  structures  in  Rugosa.     /.  Fac.  Sci.  Hokkaido  Univ.  (4,  geol. 

min.)  11  :  571-630,  3  pis. 
Kayser,  F.  H.  E.     1889.     Ueber  das  Devon  in  Devonshire  und  im  Boulonnais.     Neites  Jb. 

Miner.  Geol.  Paldont.,  Stuttgart,  1  :  1 79-191. 
Lang,    W.    D.     1923.     Trends   in   British   Carboniferous   Corals.     Proc.    Geol.    Ass.,    London, 

34  :  120-136,  text-figs.  15-17. 
Lang,  W.  D.  &  Smith,  S.     1935.     Cyathophyllum  caespitosum  Goldfuss,  and  other  Devonian 

corals  considered  in  a  revision  of  that  species.    Q.  Jl  geol.  Soc.  Lond.,  91 :  538-590,  pis.  35-37. 
Lang,  W.  D.,  Smith,  S.  &  Thomas,  H.  D.     1940.     Index  of  Palaeozoic  coral  genera,     vii  +  231 

pp.  London. 
Lecompte,     M.     1952.     Madreporaires    Paleozoiques.     In    Piveteau,     J.     (Ed.).     Traite     de 

Pale'ontologie,  1  :  419-538,  text-figs.  1-154  +  1-75.     Paris. 
Linnaeus,  C.     1758.     Systema  Naturae,  1  :  iv  +  823  pp.     Editio  Decima,  Reformata.     Stock- 
holm. 
Lloyd,  W.     1933.     The  geology  of  the  country  around  Torquay.     [2nd.  edn.].     Mem.  geol. 

Surv.  Engld.  &■  Wales,  London  :  1-169,  pis.  1-7. 
Lonsdale,  W.     1840.     In  Sedgwick,  A.  &  Murchison,  R.  I.     On  the  Physical  Structure  of 

Devonshire,  and  on  the  Subdivisions  and  Geological  Relations  of  its  older  stratified  Deposits, 

etc.     Trans,  geol.  Soc.  Lond.  (2nd  ser.),  5  :  697,  pi.  58. 
M'Coy,  F.     1851-55.     In  Sedgwick,  A.     A  Synopsis  of  the  Classification  of  the  British  Palaeo- 
zoic rocks,   by  the  Rev.  Adam  Sedgwick,    .  .  .  with  a  systematic  Description  of  the  British 

Palaeozoic  Fossils  in  the  Geological  Museum  of  the   University  of  Cambridge  by  Frederick 

M'Coy.     i-iv,  1-184  (1851);   i-[x],  185-406  (1852);   Introduction,  i-xcviii,  407-661  (1855); 

pis.  iA-iL,  iiA-iiD,  iiiA-iiiD.     London  &  Cambridge. 
McLaren,  D.   J.     1959.     A  revision  of  the  Devonian  coral  genus  Synaptophyllum  Simpson. 

Bull.  geol.  Surv.  Can.,  Ottawa,  48  :  15-33,  pls-  7-10. 


278  COLONIAL  PHILLIPS ASTRAEIDAE  FROM  S.E.  DEVON 

Martin,  W.     1809.     Petrificata  derbiensia;   or,  figures  and  descriptions  of  petrifactions  collected 

in  Derbyshire,     ix,  ii,  ii,  [1-102],  1-28,  pis.  1-52.     Wigan. 
Mayr,  E.,  Linsley,  E.  G.  &  Usinger,  R.  L.     1953.     Methods  and  principles  of  systematic  zoology. 

ix  +  328  pp.,  45  text-figs.  New  York. 
Middleton,    G.    V.     1959.     Devonian    Tetracorals    from    South    Devonshire,    England.     /. 

Paleont.,  Chicago,  33  :  138-160,  pi.  27. 
— ■ —  i960.     Spilitic  rocks   in   South-east   Devonshire.     Geol.   Mag.,   London,    97  :  192-207,   2 

text-figs. 
Moenke,  M.     1954.     Genus  Hexagonaria  in  the  Devonian  of  the  Holy  Cross  Mts.     Acta  geol. 

pol.  palaeont.,  Warszawa,  4  :  445-483,  2  pis.  [In  Polish  with  English  summary]. 
Moore,  R.  C,  Hill,  D.  &  Wells,  J.  W.     1956.     Glossary  of  morphological  terms  applied  to 

corals.     In  Moore,  R.  C.  (Ed.)  Treatise  on  Invertebrate  Paleontology,  F  :  245-251.     Law- 
rence, Kansas. 
Oliver,  W.  A.,  Jr.     i960.     Rugose  corals  from  reef  limestones  in    the    Lower    Devonian  of 

New  York.     /.  Paleont.,  Chicago,  34  :  59-100,  pis.  13-19. 
1964.     The  Devonian  colonial  coral  genus  Billingsastraea  and  its  earliest  known  species. 

Prof.  Pap.  U.S.  geol.  Surv.,  Washington,  483-B  :  iv  +  1-5,  2  pis. 
Orbigny,  A.  d'.     1849.     Note  sur  des  polypiers  fossiles.     12  pp.     Paris. 
1850.     Prodrome    de  Paleontologie   stratigraphique  universelle  des  animaux  mollusques  et 

rayonnis,  1  :  ix  +  394  pp.     Paris. 
Pedder,  A.  E.  H.     1964.     Two  new  genera  of  Devonian  tetracorals  from  Australia.     Proc. 

Linn.  Soc.  N.S.W.,  Sydney,  88  :  364-367,  pi.  19. 
1966.     The  Devonian  tetracoral  Haplothecia  and  new  Australian  phacellophyllids.     Proc. 

Linn.  Soc.  N.S.W.,  Sydney,  90  :  181-189,  pi.  6. 
Penecke,  K.  A.     1894.     Das  Grazer  Devon.     Jb.  K.  K.  geol.  Reichsanst.,  Wien,  43  (for  1893)  : 

567-616,  pis.  7-12. 
Phillips,  J.     1841.     Figures  and  Descriptions  of  the  Palaeozoic  Fossils  of  Cornwall,  Devon  and 

West  Somerset,     xii  +  231  pp.,  60  pis.     London. 
Quenstedt,  F.  A.  von     1851-52.     Handbuch  der  Petrefaktenkunde.     [iv]  +  1-528,  pis.  1-52 

(1851);    529-792,  pis.  53-62  (1852).     Tubingen. 
1878-81.     Petrefactenkunde    Deutschlands,     6.     Die    Rohren-    und    Sternkor alien.     1-144 

(1878);   145-624(1879);  625-912(1880);  913-1094(1881);  and  Atlas.     Leipzig. 
Roemer,  C.  F.     1852-54.     Kohlen-Periode.     In  Bronn,  H.  G.  &  Roemer,  C.  F.     1851-56. 

H.  G.  Bronn's  Lethaea  Geognostica  (Dritte  Auflage),  1,  2  :  1-788  and  Atlas.     Stuttgart. 
1883.     Lethaea  Geognostica,   1  Theil.     Lethaea  palaeozoica,   1,   2  :  325-544,  text-figs.  62- 

126  +  Atlas.     Stuttgart. 
Roemer,    F.   A.     1843.     Die   Versteinerungen  des  Harzgebirges.     [8]  +  xx  +  1-40,  pis.   1-12. 

Hannover. 
1855.     Beitrage    zur    geologischen    Kenntniss  des  nordwestlichen  Harzgebirges.     Dritte 

Abtheilung.     Palaeontographica,  5  :  [iv]  +  1-44,  pis.  1-8. 
Rozkowska,  M.     1953.     Pachyphyllinae  et  Phillipsastraea  du  Frasnien  de  Pologne.     Palaeont. 

pol.,  Warszawa,  5  :  1-89,  8  pis. 
1956.     Pachyphyllinae  from  the  Middle  Devonian  of  the  Holy  Cross  Mts.,  Part  1.     Acta 

palaeont.  pol.,  Warszawa,  1  :  271-330,  39  text-figs. 
1957-     Considerations   on  Middle   and   Upper   Devonian   Thamnophyllidae   Soshkina  in 

Poland,  Part  2.     Acta  palaeont.  pol.,  Warszawa,  2  :  81-153,  27  text-figs. 
i960.     Blastogeny  and  individual  variations  in  tetracoral  colonies  from  the  Devonian  of 

Poland.     Acta  palaeont.  pol.,  Warszawa,  5  :  3-64,  43  text-figs. 
1965.     Marisastridae  n.  fam.  and  Marisastrum  n.  gen.  (Devonian  corals).     Acta  palaeont. 

pol.,  Warszawa,  10  :  261-266,  2  text-figs. 
Sandberger,  G.  &  Sandberger,  F.     1856.     Die   Versteinerungen  des  rheinischen  Schichten- 

systems  in  Nassau,  8-9  :  233-564  +  i-xv,  Atlas  pis.  34-41.     Wiesbaden. 
Sanford,  W.  G.     1939.     A  review  of  the  Families  of  Tetracorals.     Am.  J.  Sci.,  New  Haven, 

237  :  295-323,  text-figs.  1-10;   401-423,  text-figs.  11-16. 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  279 

Schafer,  R.     1889.     On  Phillipsastraea  d'Orb.,  with  special  reference  to  Phillipsastraea  radiata 

S.  Woodward  sp.  and  Phillipsastraea  tuberosa  M'Coy  sp.     Geol.  Mag.,  London,  6  :  398-409, 

pi.  12. 
Schluter,  C.     1881.     Ueber  einige  Anthozoen  des  Devon.     Z.  dt.  geol.  Ges.,  Berlin,  33  :  75-108, 

pis.  6-13. 
Schouppe,   A.   von     1949.     Die   "  Thamnophyllen  "   und  ihre   Beziehungen  zur  Gruppe  des 

"  Cyathophyllum  caespitosum  ".     Palaeontographica,  Stuttgart  (A)  47  :  99-180,  pis.  9-14. 
1956.     Neue  Betrachtungen  zur  Skelettbildung  der  Pterocorallia.     Paldont.  Z.,  Berlin, 

30  :  148-162,  5  text-figs. 
1958.     Revision  des  Formenkreises  um  Phillipsastraea  d'Orb.,    "  P  achy phy  Hum"  E.   & 

H.,  Macgeea  (Webst.),  "  Thamnophylhim  "  Pen.,  Peneckiella  Soshk.  und  verwandter  Formen. 

Neues  Jb.  Geol.  Paldont.  Abh.,  Stuttgart,  106  :  139-244,  pis.  5,  6. 
Schouppe,  A.  von  &  Stacul,  P.     1963.     Bau  und  taxionomische  Fassung  des  devonischen 

Korallen-Genus  Macgeea  Webster  1889  und  seine  Beziehungen  zu  Thamnophyllum  Penecke 

1894.     Senckenburg.  leth.,  Frankfurt  a.M.,  44  :  265-299,  pi.  41. 
Schweigger,    A.    F.     1819.     Beobachtungen    auf   Naturhistorischen    Reisen.     xii  +  127    pp., 

8  pis.  Berlin. 
Scrutton,  C.  T.     1965.     The  ages  of  some  coral  faunas  in  the  Torquay  area.     Proc.  Usshev 

Soc,  Redruth,  1  :  186-188,  3  text-figs. 
1967.     Marisastridae    (Rugosa)    from    south-east    Devonshire,    England.     Palaeontology, 

London,  10  :  266-279,  pis.  40-43. 
Semenoff-Tian-Chansky,  P.     1961.     Madr£poraires  paldozoi'ques.     In  Semenoff-Tian-Chan- 

sky,  P.,  Lafuste,  J.  &  Delga,  M.  D.     Madreporaires  du  Devonien  du  Ch6noua  (Algerie). 

Bull.  Soc.  geol.  Fr.,  Paris  (7)  3  :  292-313,  pi.  9. 
Simpson,  G.  G.,  Roe,  A.  &  Lewontin,  R.  C.     i960.     Quantitative  zoology,     vii  -f-  440  pp.,  64 

text-figs.     New  York. 
Smith,  S.     1917.     Aulina  Rotiformis,  gen.  et  sp.  nov.  Phillipsastraea  Hennahi  (Lonsdale),  and 

Orionastraea,  gen.  nov.     Q.  J I  geol.  Soc.  Lond.,  72  :  280-307,  pis.  22-24. 
1945-     Upper  Devonian  corals  of  the  Mackenzie  River  region,  Canada.     Spec.  Pap.  geol. 

Soc.  Am.,  New  York,  59  :  viii  +  1-126,  pis.  1-35. 
Soshkina,    E.    D.     1939.     Verkhnedevonskie   korally    Rugosa   Urala.     Trudy  paleont.    Inst., 

Leningrad,  9  :  1-88,  pis.  1-14.     [In  Russian  with  English  summary]. 
■ 1949.     Devonskie  korally  Rugosa  Urala.     Trudy  paleont.  Inst.,  Leningrad,   15  :  1-160, 

pis.  1-58  [In  Russian]. 
1951.     Pozdnedevonskie  korally  Rugosa,  ikh  sistematika  i  evolyutsiya.      Trudy  paleont. 

Inst.,  Leningrad,  34  :  1-122,  pis.  1-24.     [In  Russian]. 
1952.     Opredelitel    devonskikh    chetyrekhluchevykh    korallov.     Trudy    paleont.     Inst., 

Leningrad,  39  :  1-127,  pis.  1-49.     [In  Russian]. 
1953-     Tsenogeneticheskie  prisposobleniya  u   nekotorykh  devonskikh  korallov   Rugosa. 

Ezheg.  vses.  paleont.  Obshch.,  Moscow,  14  :  80-90,  text-figs.  1-14.     [In  Russian]. 
— —  1954.     Devonski  chetyrekhluchevye  korally  Russkoi  Platformy.     Trudy  paleont.  Inst., 

Leningrad,  52  :  1-76,  pis.  1-19.     [In  Russian]. 
1962.     Podotryad  Cystiphyllina.     In  Orlov,  U.  A.  (Ed.)  Osnovy  paleontologii,  2.     Gubki, 

Arkheotsiaty,  Kishechnopolostnye,  Chervi  :  306-311,  pis.  1,  2  and  pi.  3,  figs.  1,  2.     Moscow. 

[In  Russian]. 
Soshkina,   E.   D.   &  Dobrolubova,   T.  A.     1962.     Otryad  Evenkiellida.      In  Orlov,   U.  A. 

(Ed.)  Osnovy  palaeontologii,  2.     Gubki,  Arkheotsiaty,  Kishechnopolostnye,  Chervi  :  333-339, 

pis.  17-21.     Moscow.     [In  Russian]. 
Spassky,  N.  Ya.     i960.     Paleontologicheskoe  obosnovanie  stratigrafii  paleozoya  Rudnogo  Altaya; 

vypusk  3.     Devonskie  chetyrekhluchevye  Korally  Rudnogo  Altaya.     143  pp.,  35  pis.    Moscow. 

[In  Russian]. 
Stephenson,  T.  A.     193 1.     Development  and  the  formation  of  colonies  in  Pocillopora  and 

Pontes.     Brit.  Mus.  (Nat.  Hist.)  Great  Barrier  Reef  Exped.  Sci.  Rept.,  London,  3,  3  :  1 13-134, 

pis.  1-6. 


280  COLONIAL  PHI  LLIPS  ASTR  AEI  D  AE  FROM  S.E.  DEVON 

Strusz,  D.  L.     1965.     Disphyllidae  and  Phacellophyllidae  from  the  Devonian  Garra  Formation 

of  New  South  Wales.     Palaeontology,  London,  8  :  518-571,  pis.  72-78. 
Stumm,  E.  C.     1949.     Revision  of  the  families  and  genera  of  the  Devonian  tetracorals.     Mem. 

geol.  Soc.  Am.,  New  York,  40  :  viii  +  1-92,  pis.  1-25. 
Taylor,   P.   W.     1951.     The   Plymouth   Limestone.     Trans.   R.  geol.   Soc.   Corn.,   Penzance, 

18  :  146-214,  pis.  1-5. 
Torley,  K.     1934.     Die  Brachiopoden  des  Massenkalks  der  oberen  Givet-Stufe  von  Bilvering- 

sen  bei  Iserlohn.     Abh.  Senckenb.  naturforsch.  Ges.,  Frankfurt  a.M.,  43,  3:  67-148,  pis.  1-9. 
Ussher,  W.  A.  E.     1903.     The  geology  of  the  country  around  Torquay.     Mem.  geol.  Surv. 

Engld.  &•  Wales,  London  :  i-iv  +  1-142,  17  text-figs. 
Ussher,  W.  A.  E.  et  al.     1913.     The  geology  of  the  country  around  Newton  Abbot.     Mem. 

geol.  Surv.  Engld.  cS-  Wales,  London  :  i-vi  -f  1-149,  3  pis. 
Verneuil,  E.  de  &  Haime,  J.     1850.     In  Verneuil,  E.  de.     Note  sur  les  fossiles  d6voniens 

du  district  de  Sabero  (Leon).     Bull.  Soc.  geol.  Fr.,  Paris  (2)  7  :  161-162. 
Voynovskiy-Kriger,  K.  G.     1954.     O  dinamike  razvitiya  septalnogo  apparata  v  ontogeneze 

chetyrekhluchevykh.     Byull.   mosk.   Obshch.   Ispyt.  Prir.,   Moscow,    59,    otdel,    geol.,    29, 

5  :  51-64,  text-figs.  1-11.     [In  Russian]. 
Walther,  C.     1929.     Untersuchungen  iiber  die  Mitteldevon-  Oberdevongrenze.     Z.  dt.  geol. 

Ges.,  Berlin,  80  (for  1928)  :  97-152,  text-figs.  1-34. 
Wang,  H.  C.     1950.     A  revision  of  the  Zoantharia  Rugosa  in  the  light  of  their  minute  skeletal 

structures.     Phil.  Trans.  R.  Soc,  London  (B)  234  :  175-246,  pis.  4-9. 
Webby,  B.  D.     1964.     Devonian  corals  and  brachiopods  from  the  Brendon  Hills,  west  Somerset. 

Palaeontology,  London,  7  :  1-22,  pi.  1. 
Webster,  C.  L.     1889.     Description  of  a  New  Genus  of  Corals,  from  the  Devonian  Rocks  of 

Iowa.     Am.  Nat.,  Lancaster,  Pa.,  23,  272  :  710-712. 
Wedekind,  R.     1922.     Zur  Kenntnis  der  Stringophyllen  des  oberen  Mitteldevon.     Sber.  Ges. 

Beford.  ges.  Naturw.  Marburg,  1921,  1  :  1-16,  text-figs.  1-18. 
Whidborne,  G.  F.     1888-1907.     A  monograph  of  the  Devonian  fauna  of  the  south  of  England, 

Part  1  :  1-344,  pis.  1— 31;   Part  2  :  1-222,  pis  1-24;   Part  3  :  1-247,  pls-  I_38-     Palaeontogr. 

Soc.  [Monogr.],  London. 

ADDENDUM 

Whilst  this  paper  was  in  press,  J.  E.  Sorauf,  (1967,  Paleont.  Contr.  Univ.  Kans., 
16  :  1-41)  published  a  work  describing  phillipsastraeids  from  the  Frasnian  of  Belgium. 
The  opportunity  is  taken  here  to  comment  briefly  on  some  of  the  more  important 
points  discussed  by  Sorauf  which  bear  closely  on  the  present  paper. 

Sorauf  (pp.  5,  15)  introduces  the  term  "  pseudocerioid  ",  which  he  defines  and 
uses  in  precisely  the  same  way  as  in  the  present  paper.  It  is  gratifying  to  note  that 
independent  work  by  Sorauf  and  the  writer  on  the  same  species  (principally  Frechas- 
traea  goldfussi  and  F.  pentagona)  has  led  to  the  same  interpretation  of  the  corallite 
wall  structure.  Sorauf  (p.  5),  however,  infers  that  species  of  Phillipsastrea  and  the 
Phillipsastraeidae  are  never  cerioid  whereas  the  genus  and  family  as  defined  here  do 
include  massive  forms  with  an  epitheca  between  at  least  some  of  the  corallites 
(this  paper,  p.  210). 

Sorauf  (p.  13)  separates  Phillipsastrea  and  Pachyphyllum  by  restricting  the  latter 
genus  to  species  with  a  perfect  single  series  of  horseshoe  dissepiments.  In  Phillipsas- 
trea he  includes  a  complete  range  of  dissepimental  form,  from  specimens  lacking  any 
sign  of  horseshoe  dissepiments  to  those  in  which  an  almost  complete  series  is  present. 
In  fact  this  whole  range  is  represented  in  Sorauf's  concept  of  a  single  species  of 
Phillipsastrea,  P.  hennahi  (fig.  5,  xc  and  2). 


COLONIAL  PHILLIPSASTRAEIDAE  FROM  S.E.  DEVON  281 

Apart  from  the  difference  in  the  degree  of  development  of  the  horseshoe  dissepi- 
ments, species  of  Phillipsastrea  and  Pachyphyllum  show  no  significant  divergence  in 
their  basic  morphology  to  warrant  generic  separation.  Moreover,  when  the  develop- 
ment of  horseshoe  dissepiments  is  closely  examined  even  this  criterion  is,  in  the 
writer's  opinion,  impossible  to  maintain  (compare  the  longitudinal-section  of  a 
topotype  of  Pachyphyllum  bouchardi  (Semenoff-Tian-Chansky  1961,  pi.  9,  fig.  2)  with 
a  section  attributed  by  Sorauf  (fig.  5,  2)  to  Phillipsastrea  hennahi). 

The  writer  is  also  unable  to  agree  with  Sorauf 's  (p.  26)  interpretation  of  Phillipsas- 
trea hennahi.  None  of  the  illustrations  (figs.  5,  la-c;  8,  la-d:  figs.  5,  id-e;  5,  2 
are  longitudinal-sections  only)  is  considered  consubspecific  with  Phillipsastrea 
hennahi  hennahi  herein,  whilst  two  specimens  (figs.  5,  2;  8,  la-b)  are  possibly  con- 
subspecific with  P.  hennahi  ussheri  subsp.  nov.  The  significance  of  this  should  not 
be  overlooked.  P.  hennahi  hennahi  appears  to  be  characteristic  of  the  middle 
and  upper  Givetian  of  England  (this  paper,  p.  216),  not  of  the  Frasnian  as  stated 
by  Sorauf  (pp.  23,  27).  On  the  other  hand,  the  subspecies  P.  hennahi  ussheri  is  found 
in  the  English  Lower  Frasnian. 

The  same  English  Lower  Frasnianlimestones  yield  Frechastraeapentagonapentagona 
and  F.  pentagona  minima.  Both  subspecies  were  considered,  the  latter  by  inference, 
to  be  upper  Frasnian  index  forms  by  Sorauf  (pp.  31,  33). 


geol.  15,  5  27 


PLATE  i 
Phillipsastrea  hennahi  hennahi  (Lonsdale) 

Fig.  i.     Cross-section  (slide).     GSM  PFi 245  (cut  from  lectotype).      X3. 

Figs.  2,  3.     Longitudinal-sections  (slide) .     GSM  PFi 249  (cut  from  lectotype).      X4. 

Fig.  4.     Longitudinal-section  (peel).     GSM  PF4028  (taken  from  lectotype).      X4. 
Figs.  1-4  all  upper  Givetian,  Barton  Quarry. 

Fig.  5.     Longitudinal-section  (peel).       TM(JB)  79;    upper  Givetian,  Lummaton  Quarry. 
X4. 

Fig.  6.     Longitudinal-section  (peel).    GSM  PF4028  (taken  from  lectotype) ;  upper  Givetian, 
Barton  Quarry.      X4. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  1 


GEOL.  15,  5 


28 


PLATE  2 
Phillipsastrea  hennahi  hennahi  (Lonsdale) 

Fig.  i.      Cross-section  (peel).     TM(JB)  79;    upper  Givetian,  Lummaton  Quarry.      X3. 

Fig.  2,  3.  Cross-  and  longitudinal-sections  (peels).  OUM  D74/PI  (from  holotype  of  Astraea 
inter cellulosa  Phillips) ;    PMiddle  Devonian,  Torquay.      X2. 

Fig.  4.  Cross-section  (peel).  GSM  PF4029  (from  lectotype  of  Syringophyllum  cantabricum 
Edwards  &  Haime) ;    PMiddle  Devonian,  Torquay.      X2. 


Bull.  Br.  Mus.  nal.  Hist.  (Geol.)  15,  5 


PLATE  2 


GEOL.   15,  5 


PLATE  3 

Phillipsastrea  hennahi  ussheri  subsp.  nov. 

Fig.  i.     Cross-section  (slide).     OUM  D544/P  i.      X3. 

Fig.  2.     Longitudinal-section  (slide).     OUM  D544/P2.      X5. 

Both  cut  from  holotype;  Lower  Frasnian,  road  cutting  20  yd.  west  of  Ramsleigh  Quarry 

entrance. 
Fig.  3.     Cross-section    (peel).       BM(NH)    R5616;     Lower    Frasnian,    Ramsleigh    Quarry. 
X3- 


Bull.  Br.  Mus.  rial.  Hist.  (Geol.)  15,  5 


PLATE  3 


PLATE  4 

Phillipsastrea  devoniensis  (Edwards  &  Haime) 

Fig.  i.     Cross-section  (peel).     OUM  D277/PI ;  PMiddle  Devonian,  Rocky  Valley,  Torquay. 
X2. 
Fig.  2.     Cross-section  (slide).     BM(NH)  R29996;    Middle  Devonian,  Torquay.      X2. 
Fig.  3.     Longitudinal-section  (peel).      X3. 
Fig.  4.     Cross-section  (peel).      X2. 

Figs.  3  and  4  both  TM(JB)  105;    (?)  upper  Givetian,  Lummaton  Quarry. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,    5 


PLATE  4 


"gggpsew 


PLATE  5 

Phillipsastrea  ananas  (Goldfuss) 

Fig.  i.     Longitudinal-section  (slide).     BM(NH)  R46158C.      X4. 
Fig.  2.     Cross-section  (slide).     BM(NH)  R.46i58a.      X2. 
Fig.  3.     Cross-section  (slide).     BM(NH)  R.46i5ga.      X2. 
Fig.  4.     Longitudinal-section  (slide).     BM(NH)  R461596.      X4. 
All  Lower  Frasnian,  road  cutting  south  side  30-35  yds.  west  of  Ramsleigh  Quarry  entrance. 


Bull.  Br.  Mus.  not.  Hist.  (Geol.j  15,  5 


PLATE  5 


PLATE  6 

Phillipsastrea  rozkowskae  sp.  nov. 

Fig.  i.     Cross-section  (slide).     BM(NH)  R461 56a  (cut  from  holotype).      X2-5. 
Figs.  2,  3.     Longitudinal-sections  (slide).     BM(NH)  R46156&  (cut  from  holotype).      X5. 
Fig.  4.     Cross-sections  (slide).     BM(NH)  R4615712.      X2-5. 
All  Lower  Frasnian,  road  cutting  south  side  25  yds.  west  of  Ramsleigh  Quarry  entrance. 

Frechastraea  pentagona  pentagona  (Goldfuss) 

Fig.  5.     Cross-section  (slide).      OUM  D537/PI ;   Lower  Frasnian,  road  cutting  20  yds.  west 
of  Ramsleigh  Quarry  entrance.      X4. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  6 


PLATE  7 
Frechastraea  pentagona  pentagona  (Goldfuss) 

Fig.  i.  Cross-section  (slide).  OUM  D537/PI ;  Lower  Frasnian,  road  cutting  20  yds. 
west  of  Ramsleigh  Quarry  entrance.      X  8. 

Fig.  2.     Cross-section  (peel).      OUM  D279/PI ;    Lower  Frasnian,  Ramsleigh  Quarry.    x6. 

Figs.  3,  4.  Longitudinal-sections  (slide).  OUM  D537/P2;  Lower  Frasnian,  road  cutting 
20  yds.  west  of  Ramsleigh  Quarry  entrance.      x8. 

Fig.  5.  Longitudinal-section  (slide).  Geol.-Pal.  Inst.  Bonn,  Goldfuss  Colin.  206  (cut  from 
lectotype);   Frasnian,  Namur,  Belgium.      x8. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  7 


PLATE  8 
Frechastraea  pentagona  (Goldfuss)  minima  (Rozkowska) 

Fig.  i.  Cross-section  (peel).  GSM  PF4031  (taken  from  GSM  73118);  Lower  Frasnian, 
Ramsleigh  Quarry.      X4. 

Figs.  2,  3.  Longitudinal-sections  (slide).  GSM  PF4032  (cut  from  GSM  73118);  Lower 
Frasnian,  Ramsleigh  Quarry,      x  8. 

Frechastraea  micrommata  (C.  F.  Roemer) 

Figs.  4,  5.  Cross-  and  longitudinal-sections  (slides).  Geol.-Pal.  Inst.  Bonn  34  (cut  from 
lectotype);    Frasnian,  Ferques  near  Boulogne,  France.      X4. 


Bull.  By.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  8 


PLATE  9 
Frechastraea  carinata  sp.  nov. 

Fig.  i.  Cross-section  (slide).  OUM  D3096  (cut  from  holotype) ;  Lower  Frasnian,  road 
cutting  80  yds.  west  of  Ramsleigh  Quarry  entrance.      x  6. 

Fig.  2.     Cross-section  (peel).     BM(NH)  R5634;   Lower  Frasnian,  Ramsleigh  Quarry.      x6. 

Fig.  3.  Longitudinal-section  (slide) .  OUM  D^ogd  (cut  from  holotype) ;  Lower  Frasnian, 
road  cutting  80  yds.  west  of  Ramsleigh  Quarry  entrance.      x  6. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  9 


PLATE  10 

Frechastraea  goldfussi  (de  Verneuil  &  Haime) 

Fig.  i.     Cross-section  (slide).     OUM  D540/P2.      X5. 
Figs.  2,  3.     Cross-sections  (slide).     OUM  D539/P2.      X5. 
Fig.  4.     Longitudinal-section  (slide).     OUM  D541/P2.      x6. 
Fig.  5.     Longitudinal-section  (slide).     OUM  D540/PI.      x8. 

All  Lower  Frasnian,  Ramsleigh  Quarry. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  10 


PLATE  ii 

Frechastraea  goldfussi  (de  Verneuil  &  Haime) 

Fig.  i.     Longitudinal-section  (slide) .     OUM  D540/PI.      x8. 
Fig.  2.     Cross-section  (peel).     TM(JB)  318.      X7. 

Figs.  1  and  2  both  Lower  Frasnian,  Ramsleigh  Quarry. 

Frechastraea  bowerbanki  (Edwards  &  Haime) 
Fig.  3.     Cross-section  (peel).     TM  136/7;   Lower  Frasnian,  Ramsleigh  Quarry.      xio. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  11 


geol.  ig,  5 


29 


PLATE  12 

Frechastraea  bowerbanki  (Edwards  &  Haime) 

Fig.  i.     Cross-section  (peel).     TM  136/7.      X3. 

Fig.  2.     Longitudinal-section  (slide).     BM(NH)  R46373.      X6. 

Fig.  3.     Longitudinal-section  (slide).     TM  136/7.      x6. 

All  Lower  Frasnian,  Ramsleigh  Quarry. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  12 


PLATE  13 

Thamnophyllum  gertnanicum  schouppei  nom.  nov. 

Fig.  1.     Cross-section  (slide).     OUM  D509/PI ;    lower  Givetian,  Dyer's  Quarry.      X50. 
Fig.  2.     Cross-section  (slide).      Pal.   Inst.  Graz  UPG  327   (cut  from  holotype);    Middle 
Devonian,  Torquay.      X4. 

Fig.  3.     Cross-section  (peel).     OUM  D507/3/PI.      X4. 
Fig.  4.     Cross-section  (slide).     OUM  D509/PI.  X4. 

Figs.  3  and  4  both  lower  Givetian,  Dyer's  Quarry. 


Bull.  Br.  Mus.  nat  Hist.  (Geol.)  15,  5 


PLATE  13 


PLATE  14 

Thamnophyllum  germanicum  schouppei  nom.  nov. 

Figs.  1,  2.     Longitudinal-  and  cross-sections  (slide).     OUM  D509/PI.      X4. 
Fig.  3.     Longitudinal-section  (peel).     OUM  D507/2/PI.      X4. 
Fig.  4.     Longitudinal-section  (peel) .     OUM  D508/4/P2.      X4. 
All  lower  Givetian,  Dyer's  Quarry. 


Bull.  Br.  Mus.  nat   Hut.  (Geol.)  15,  5 


Fig. 

i. 

X4- 

Fig. 

2. 

Fig. 

3- 

Fig. 

4- 

PLATE  15 
Thamnophyllum  caespitosum  paucitabulatum  subsp.  nov. 
Cross-section   (slide).     BM(NH)   R46163&;    upper  Givetian,  Lummaton  Quarry. 

Thamnophyllum  caespitosum  (Goldfuss)  sensu  lato 

Longitudinal-section  (slide).     BM(NH)  R.46i68a.      X4. 

Cross- and  longitudinal-sections  (slide).     BM(NH)  R46i7ia.      X4. 

Cross-section  (slide).     BM(NH)  R.46i68a.      X4. 

Figs.  2-4  all  middle  Givetian,  Wolborough  Quarry. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  15 


PLATE  16 

Thamnophyllum  caespitosum  (Goldfuss)  sensu  lato 

Fig.  I.     Cross- and  longitudinal-sections  (slide).     BM(NH)  1146168a.      xi. 
Fig.  2.     Longitudinal-section  (slide).     BM(NH)  R46175&.      X4. 

Both  middle  Givetian,  Wolborough  Quarry. 

Thamnophyllum  caespitosum  paucitabulatum  subsp.  nov. 

Fig.  3.     Cross-section  (slide).     BM(NH)  R46163&.      xi. 

Fig.  4.     Cross-section  (slide).     BM(NH)  R46i65«  (cut  from  holotype).      XI. 
Fig.  5.     Cross-section  (slide).     BM(NH)  ^6165^  (cut  from  holotype).      X4. 
Fig.  6.     Longitudinal-section  (slide) .     BM(NH)  R461 65c  (cut  from  holotype).      X4. 
Fig.  7.     Longitudinal-section  (slide) .     BM(NH)  R461 656  (cut  from  holotype).      X4. 
Figs.  3-7  all  upper  Givetian,  Lummaton  Quarry. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15.  5 


PLATE  16 


PLATE  17 

Thamnophyllum  caespitosum  paucitabulatum  subsp.  nov. 

Fig.  1.     Cross-section  (slide).     BM(NH)  R461 65a  (cut  from  holotype).      X4. 
Fig.  2.     Longitudinal-section  (slide).     BM(NH)  R46163C.      X4. 
Fig.  3.     Longitudinal-section  (slide).     BM(NH)  R46i64^.      X4. 
All  upper  Givetian,  Lummaton  Quarry. 

Peneckiella  minor  minor  (F.  A.  Roemer) 

Fig.  4.     Cross -section  (slide).      X3. 
Fig.  5.     Longitudinal -section  (slide).      x6. 
Figs.  4  and  5  both  Bergakad.  Clausthal-Z.  117  (cut  from  holotype);    Frasnian,  Winterburg 

near  Bad  Grund,  Germany. 


Bull.  Br.  Mus.  nat.  Hist.  (Geol.)  15,  5 


PLATE  17 


PLATE  18 

Peneckiella  salternensis  sp.  nov. 

Fig.  i.     Cross-section  (slide).     OUM  D546/P1.      X3. 

Fig.  2.     Cross-section  (slide) .     OUM  D548/PI.  (cut  from  holotype).      X3. 

Fig.  3.     Longitudinal -section  (slide).     OUM  D548/P3  (cut  from  holotype).      X4. 

Fig.  4.     Longitudinal-section  (slide) .     OUM  D546/P3.      X4. 

All  Frasnian,  main  Peneckiella  horizon,  Saltern  Cove. 


Bull.  Br.  Mus.  nat.  Hist.  (Zool.)  15,  5 


PLATE  if 


PRINTED  IN  GREAT  BRITAIN 
BY  ADLARD  &  SON  LIMITED 
BARTHOLOMEW   PRESS,   DORKING 


->  MAY  1966 


SOME  STROPHOMENACEAN    ^lS 

BRACHIOPODS 

FROM  THE  BRITISH  LOWER 

SILURIAN 


L.  R.  M.  COCKS 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.  15  No.  6 

LONDON:   1968 


SOME   STROPHOMENACEAN  BRACHIOPODI 
FROM  THE  BRITISH  LOWER  SILURIAN 


BY 

LEONARD  ROBERT  MORRISON  COCKS 

British  Museum  (Natural  History) 


A 


Pp.  283-324;  14  Plates;  1  Text-figure 


BULLETIN  OF 
THE  BRITISH  MUSEUM  (NATURAL  HISTORY) 
GEOLOGY  Vol.   15  No.  6 

LONDON:   1968 


THE  BULLETIN  OF  THE  BRITISH  MUSEUM 
(NATURAL  HISTORY),  instituted  in  1949,  is 
issued  in  five  series  corresponding  to  the  Departments 
of  the  Museum,  and  an  Historical  series. 

Parts  will  appear  at  irregular  intervals  as  they  become 
ready.  Volumes  will  contain  about  three  or  four 
hundred  pages,  and  will  not  necessarily  be  completed 
within  one  calendar  year. 

In  1965  a  separate  supplementary  series  of  longer 
papers  was  instituted,  numbered  serially  for  each 
Department. 

This  paper  is  Vol.  15,  No.  6  of  the  Geological  (Palae- 
ontological)  series.  The  abbreviated  titles  of  periodicals 
cited  follow  those  of  the  World  List  of  Scientific  Periodicals . 


World  List  abbreviation : 
Bull.  Br.  Mus.  nat.  Hist.  (Geol. 


Trustees  of  the  British  Museum  (Natural  History)  196S 


TRUSTEES    OF 
THE     BRITISH    MUSEUM     (NATURAL    HISTORY) 

Issued  14  May  1968  Price  £2     10s 


SOME   STROPHOMENACEAN  BRACHIOPODS 
FROM  THE  BRITISH  LOWER  SILURIAN 

By  L.  R.  M.  COCKS 


CONTENTS 

I.     Introduction  .... 

II.     Systematic  descriptions  . 

Superfamily  Strophomenacea  King  . 
Family  Strophomenidae  King 
Subfamily  Strophomeninae  King 
Genus  Pentlandina  Bancroft  . 

Pentlandina  tarlana  Bancroft 
Pentlandina  pavva  Bancroft 
Pentlandina  parabola  sp.  nov. 
Pentlandina  sp. 
Subfamily  Furcitellinae  Williams 
Genus  Katastrophomena  nov. 

Katastrophomena  woodlandensis  (Reed) 
Katastrophomena  scotica  (Bancroft)   . 
Katastrophomena  penkillensis  (Reed) 
Katastrophomena  sp. 
Subfamily  Leptaeninae  Hall  &  Clarke 
Genus  Leptaena  Dalman 

Leptaena  martinensis  sp.  nov.  . 
Leptaena  havevfordensis  Bancroft 
Leptaena  valida  Bancroft 
Leptaena  urbana  Bancroft 
Leptaena  contermina  sp.  nov.    . 
Leptaena  valentia  sp.  nov. 
Leptaena  valentia  mullochensis  subsp.  nov 
Leptaena  zeta  Lamont 
Leptaena  reedi  sp.  nov.    . 
Leptaena  ziegleri  sp.  nov. 
Leptaena  quadrata  Bancroft 
Leptaena  purpurea  sp.  nov. 
Leptaena  sp.  .... 

Genus  Cvphomena  Cooper 

Subgenus  Cyphomenoidea  nov. 

C.  (Cyphomenoidea)  wisgoriensis  (Lamont 
Subgenus  Laevicyphomena  nov.    . 

C.  (Laevicvphomena)  feliciter  sp.  nov. 
Genus  M ackerrovia  gen.  nov. 

Mackerrovia  lobatus  (Lamont  &  Gilbert) 
Genus  Bellimurina  Cooper 

Bellimurina  sp.        .... 
Acknowledgments  ..... 

References     ...... 


III. 
IV. 

GEOL.    15,   0. 


&  Gilbert 


286 

286 
286 
288 
289 
289 
289 

291 
292 

293 
293 
293 
295 

296 

297 
298 
299 
299 
302 
304 
305 
305 
306 

307 
309 
309 
310 
311 

312 

313 
314 
315 
316 
316 
317 
317 
319 

3*9 

321 
321 
322 
322 
3° 


286  BRITISH  LOWER  SILURIAN 

SYNOPSIS 
The  brachiopod  family  Leptaenidae  is  relegated  to  subfamilial  rank  within  the  Strophomenidae. 
All  the  known  British  Lower  Silurian  (Llandovery)  species  from  the  revised  family  Stropho- 
menidae are  described  and  figured.  Two  new  genera  are  erected :  Katastrophomena,  type 
species  Strophomena  antiquata  var.  woodlandensis  Reed  1917,  and  Mackevrovia,  type  species 
Brachy 'prion  arenaceus  var.  lobatus  Lamont  &  Gilbert  1945.  Two  new  subgenera  are  erected 
within  the  genus  Cyphomena  Cooper  1956  (hitherto  thought  to  be  confined  to  the  Ordovician) : 
Cvphomenoidea,  type  species  Leptaena  wisgoriensis  Lamont  &  Gilbert  1945,  and  Laevicyphomena, 
type  species  C.  (L.)  feliciter  sp.  nov.  Eight  new  species  and  one  subspecies  are  erected,  and 
the  ecological  communities  of  all  the  species  recorded. 

I.     INTRODUCTION 

Three  families  of  the  superfamily  Strophomenacea  occur  in  the  British  Lower 
Silurian.  One,  the  Stropheodontidae,  has  already  been  considered  (Cocks  1967), 
and  thus  the  present  paper  is  concerned  exclusively  with  representatives  of  the 
other  two  families,  the  Strophomenidae  and  the  Leptaenidae. 

Strophomenids  are  widely  distributed  in  the  Lower  Silurian  of  Britain,  but  are 
sporadic  and  rare  at  most  localities.  They  have  been  recorded  under  the  name 
'  Strophomena  antiquata  '  in  most  faunal  lists.  Leptaenids  are  rather  commoner  and 
have  usually  been  referred  to  '  Leptaena  rhomboidalis  '.  Although  the  Wenlock 
species  of  both  families  have  been  figured  for  many  years,  notably  by  Thomas 
Davidson  in  various  works,  the  Llandovery  species  have  remained  for  the  most 
part  poorly  described. 

The  stratigraphy  and  correlation  of  the  British  Llandovery  is  at  the  present 
time  under  review,  and  a  joint  paper  by  Dr.  A.  M.  Ziegler,  Dr.  W.  S.  McKerrow 
and  the  present  author  is  in  course  of  preparation.  Correlation  between  the  various 
areas  and  the  type  area  of  Llandovery  itself  has  been  effected  mainly  by  the  use  of 
evolving  brachiopod  lineages  such  as  those  of  Stricklandia  (Williams  195 1)  and 
Eocoelia  (Ziegler  1966). 

II.     SYSTEMATIC   DESCRIPTIONS 

Superfamily  STROPHOMENACEA  King,  1846 
Classification 

In  the  recent  Treatise  (Williams  et  al.  1965),  the  classification  of  the  Stropho- 
menacea is  as  follows : 

Superfamily  Strophomenacea  King  1846  (L.Ord.-L.Carb.) 
Family  Strophomenidae  King  1846  (L.Ord.-L.Dev.) 
Subfamily  Strophomeninae  King  1846  (M.Ord.-U.Sil.) 
Furcitellinae  Williams  1965  (M.Ord.-L.Sil.) 
Rafinesquininae  Schuchert  1893  (M.-U.Ord.) 
Glyptomeninae  Williams  1965  (L.-M.Ord.) 
Oepikinae  Sokolskaya  i960  (M.-U.Ord.) 
Leptaenoideinae  Williams  1953  (U.Sil.-L.Dev.) 
Family  Foliomenidae  Williams  1965  (U.Ord.) 
Family  Christianiidae  Williams  1953  (Ord.) 
Family  Leptaenidae  Hall  &  Clarke  1894  (M.Ord.-L.Carb.) 
Family  Stropheodontidae  Caster  1939  (U.Ord.-U.Dev.) 


STROPHO  M  E  N  A  CEANBRACHIOPODS  287 

The  Stropheodontidae  is  divided  into  seven  subfamilies  which  are  outside  the 
scope  of  the  present  paper;  the  family  is  distinct  in  possessing  denticles,  and  was 
held  with  some  justification  to  possess  superfamilial  status  by  Sokolskaya  (i960  : 
213).  Of  the  remaining  families,  the  Foliomenidae  and  the  Christianiidae  are 
confined  to  the  Ordovician,  leaving  the  Strophomenidae  and  Leptaenidae  to  be 
considered  here. 

There  is  some  doubt  as  to  whether  these  two  families  are  validly  separable.  Apart 
from  the  undiscussed  list  at  the  end  of  Hall  &  Clarke  (1894  :  353-354),  which  divided 
what  are  now  known  as  the  suborders  Strophomenidina  and  Triplesiidina  into  a 
rather  arbitrary  division  between  Strophomenidinae  and  Leptaenidae  (thus  creating 
the  latter  family),  the  two  families  were  not  divided  again  until  1956  in  Cooper's 
great  work  on  the  Chazyan  brachiopods.  Thus  Williams  in  his  paper  on  stropho- 
menoid  classification  (1953)  did  not  mention  the  Leptaenidae:  he  divided  the 
Strophomenacea  into  three  families,  the  Strophomenidae,  Stropheodontidae  and 
Christianiidae,  and  the  Strophomenidae  was  divided  into  only  two  subfamilies, 
the  Strophomeninae  and  the  Leptaenoideinae.  In  fact  he  specifically  cited  the 
Rafinesquinidae  and  Strophomenidae  as  synonymous  (1953  :  8). 

Cooper  (1956)  divided  the  Chazyan  Strophomenacea  into  three  families,  Leptaen- 
idae, Christianiidae  and  Strophomenidae.  He  briefly  defined  the  three  families  as 
follows — Leptaenidae:  '  Compressed  to  faintly  lenticular  Strophomenacea  with 
large  apical  foramen'  (1956:820);  Christianiidae:  'Smooth  or  finely  costellate 
Strophomenacea  having  4  prominent  septa  in  the  brachial  valve  '  (1956  :  859) ; 
and  Strophomenidae :  '  Strophomenacea  having  either  normal  or  reversed  convexity 
of  the  valves  and  a  small  foramen  in  the  pedicle  valve  '  (1956  :  866).  Thus,  by 
those  definitions,  the  only  difference  between  a  leptaenid  and  a  strophomenid  of 
normal  convexity  lies  in  the  size  of  the  foramen.  As  many  species  of  Leptaena 
itself,  e.g.  L.  salopiensis  Williams,  possess  a  foramen  which  is  '  small,  commonly 
sealed  in  adult  shells  '  (Williams  1963  :  461),  this  definition  cannot  be  used  in  the 
type  genus  of  the  family.  Even  the  geniculation  and  disc  rugae  may  be  seen  on 
some  genera,  for  example  Luhaia  Roomusoks  1956,  ascribed  to  the  Strophomeninae 
by  Williams  (1965  :  H384). 

However,  there  does  occur  a  group  of  strophomenaceans  which  are  normally 
geniculate  and  often  possess  rugae  over  most  of  the  disc,  and  which  may  conveniently 
be  grouped  with  Leptaena  itself.  But,  bearing  in  mind  the  wide  differences  between 
the  Strophomenidae,  Christianiidae  and  Stropheodontidae,  familial  recognition  as 
the  Leptaenidae  seems  an  unwarrantedly  high  taxonomic  rank.  In  this  paper 
they  will  be  treated  as  a  subfamily  within  the  Strophomenidae,  the  Leptaeninae, 
although  even  this  separation  may  not  be  valid,  and  some  of  the  subfamilies  may 
be  inter-phyletic. 

Although  only  three  are  considered  in  this  paper,  this  arrangement  leaves  a  total 
of  seven  subfamilies  within  the  Strophomenidae,  and  their  relative  phylogenies 
leave  much  scope  for  discussion.  Certainly  the  morphology  of  the  two  genera  in 
the  Leptaenoideinae,  Leptaenoidea  and  Leptaenisca,  seem  to  indicate  descent  from 
leptaenids  rather  than  from  other  strophomenids,  and  this  is  supported  by  their 
known  stratigraphical  range. 


288 


BRITISH  LOWER  SILURIAN 


Ecological  occurrence 

The  species  described  in  this  paper  are  distributed  as  follows  in  the  animal  com- 
munities established  in  the  British  Llandovery  (Ziegler  1965,  Cocks  1967a,  Ziegler, 
Cocks  &  Bambach  1968).  The  list  omits  Leptaena  urbana  Bancroft,  whose 
community  is  unknown.  No  strophomenids  have  yet  been  found  in  the  Lingitla 
community,  which  is  thought  to  have  been  the  shallowest. 


Pentlandina  tartana  Bancroft 
Pentlandina  parva  Bancroft 
Pentlandina  parabola  sp.  nov. 
Katastrophomena  woodlandensis  (Reed 
Katastrophomena  scotica  (Bancroft) 
Katastrophomena  penkillensis  (Reed) 
Leptaena  martinensis  sp.  nov. 
Leptaena  haverfordensis  Bancroft 
Leptaena  valida  Bancroft 
Leptaena  contermina  sp.  nov 
Leptaena  valentia  sp.  nov. 
Leptaena  zeta  Lamont 
Leptaena  reedi  sp.  nov. 
Leptaena  ziegleri  sp.  nov. 
Leptaena  quadrata  Bancroft 
Leptaena  purpurea  sp.  nov.  . 
C.  {Cyphomenoidea)  wisgoriensis 

(Lamont  &  Gilbert) 
C.  (Laevicvphomena)  feliciter  sp.  nov 
Mackerrovia  lobatus  (Lamont  &  Gilbert) 


Eocoelia 
Community 

Pentamerus 
Community 

Stricklandia 
Community 

Clorinda 
Community 

— 

— 

X 

X 

X 
X 
X 

X 
X 
X 





X 

X 

— 

X 
X 
X 
X 

X 

X 

? 

X 

. 

— 

X 
X 

X 

. 

■ 

X 

x 

? 

) 



X 

— 

Family  STROPHOMENIDAE  King,  1846 

Apart  from  the  Leptaeninae,  there  are  only  two  groups  of  Strophomenidae  which 
have  so  far  been  found  in  rocks  of  Llandovery  age,  as  recognized  by  Williams  (195 1  : 
115),  who  at  that  time  referred  them  provisionally  to  Strophomena  Rafmesque  and 
Holtedahlina  Foerste. 

The  two  groups  are  here  referred  to  Katastrophomena  gen.  nov.  and  Pentlandina 
Bancroft  1949.  These  are  placed  in  different  subfamilies,  the  Furcitellinae  and  the 
Strophomeninae,  and  are  the  only  post-Ordovician  genera  known  in  either  sub- 
family. The  chief  subfamilial  difference  (Williams  1965  :  H384,  H386)  is  that  the 
Strophomeninae  are  unequally  parvicostellate  and  the  Furcitellinae  are  costellate, 
but  in  fact  later  species  of  Katastrophomena  (such  as  K.  penkillensis  (Reed)  described 
below)  become  unequally  parvicostellate.  Thus  in  the  two  genera  concerned,  the 
chief  differences  in  Silurian  species  lie  in  the  internal  structures  of  the  brachial  valve, 
and  the  usual  presence  of  a  strong  fold  and  sulcus  in  Pentlandina. 


STROPHOMENACEAN  BRACHIOPODS  289 

Subfamily  STROPHOMENINAE  King,  1846 
Genus  PENTLANDINA  Bancroft,  1949 

1949.     Strophomena  (Pentlandina)  Bancroft  :  11,  13. 

1965.  Pentlandina  Bancroft  Williams  :  H384. 

1966.  Pentlandina  Bancroft;  Boucot  et  al.  :  25. 

Diagnosis  :  Biconvex  to  convexo-concave  small  stophomeninids  with  prominent 
fold  and  sulcus,  often  with  an  ornament  of  parvicostellae  interrupting  broken  rugae. 

Type  species  (by  original  designation) ;  Strophomena  {Pentlandina)  tartana  Ban- 
croft 1949  from  the  Upper  Llandovery  of  Deerhope  Burn,  Pentland  Hills,  Scotland. 

Species  assigned : 

Strophomena  (Pentlandina)  tartana  Bancroft  1949  :  13.  Upper  Llandoverv,  Pent- 
land  Hills,  Scotland. 

Strophomena  (Pentlandina)  parva  Bancroft  1949  :  13,  pi.  1,  fig.  9.  Upper  Llan- 
dovery, The  Frolic,  Haverfordwest,  Pembrokeshire. 

Pentlandina  parabola  sp.  nov.     Upper  Llandovery,  Purple  Shale,  Shropshire. 

Strophomena  hirnndo  Barrande  1879,  pi.  47,  figs.  1-32  pars.  Wenlock,  Bohemia, 
Czechoslovakia. 

Leptaena  loveni  de  Verneuil  1848  :  339,  pi.  4,  fig.  5.  Visby  Marl  (Upper  Llan- 
dovery) Gotland,  Sweden. 

Leptaena  parvula  Kindle  1915  :  14,  pi.  i,  figs.  5-9.  Stonewall  Limestone,  Sas- 
katchewan, Canada. 

Leptaena  sinuosus  Kindle  1915  :  13,  pi.  1,  figs.  1-4.  Stonewall  Limestone, 
Saskatchewan,  Canada. 

i'Leptaena  lewisii  Davidson  1847  :  59,  pi.  12,  figs.  22-24,  Lower  Wenlock,  Rushall 
Canal,  Staffordshire. 

Discussion.  Pentlandina  was  raised  to  generic  level  by  Williams  (1965)  and 
placed  within  the  Strophomeninae.  Boucot  et  al.  (1966)  recently  removed  the  genus 
to  the  Leptaenidae  on  account  of  the  pedicle  muscle  field  and  brachial  processes 
and  adductor  plates  (=  trans-muscle  septa?).  It  is,  however,  quite  impossible  to 
agree  with  this  point,  as  each  of  the  cited  structures  in  Pentlandina  is  firmly 
attributable  to  the  Strophomeninae  rather  than  to  the  Leptaeninae.  In  fact  the 
morphology  of  Pentlandina  tartana,  the  type  species  described  below,  is  not  like  any 
member  of  the  leptaeninids.  Taken  with  a  shape  so  typical  of  the  subfamily,  there 
can  be  no  doubt  that  the  genus  lies  within  the  Strophomeninae.  Gvmnarella 
Spjeldnaes  1957  has  a  similar  ornament  to  Pentlandina,  but  is  geniculate  as  opposed 
to  biconvex  and  sulcate. 

Pentlandina  tartana  Bancroft 

(PI.  1,  figs.  1-6) 

1868.     Strophomena  antiquata  (J.  de  C.  Sowerby) ;  Davidson  :  17,  pi.  2,  figs.  21-23. 

1871.     Strophomena  antiquata  (J.  de  C.  Sowerby);  Davidson  :  299  pars,  pi.  44,  figs.  7-9  only- 

1949.     Strophomena  (Pentlandina)  tartana  [Lamont  MS]  Bancroft  :  13,  non  pi.  1,  fig.  10. 


2go  BRITISH  LOWER  SILURIAN 

Diagnosis:  Small  strophomeninid  with  prominent  fold  and  sulcus.  Fine  orna- 
ment of  differentiated  parvicostellae  which  break  irregular  small  weak  rugae. 

Description.  Exterior.  Biconvex  to  slightly  resupinate  with  a  semicircular 
outline  and  small  ears.  Large  sulcus  in  pedicle  valve  with  corresponding  fold  in 
brachial  valve.  Fine  ornament  of  differentiated  parvicostellae,  and  very  weak 
irregular  rugae  of  small  wavelength  distributed  over  all  the  shell.  Interarea  of 
variable  size,  larger  in  the  pedicle  than  in  the  brachial  valve.  Large  delthyrium 
closed  at  the  apex  by  a  small  pseudodeltidium  (Plate  i,  fig.  4).  Information  un- 
certain as  to  the  chilidium,  there  is  at  least  a  small  one  developed,  but  it  is  not  clear 
whether  or  not  part  of  the  delthyrium  remained  open. 

Pedicle  interior.  Straight  hingeline  with  prominent  teeth  connected  to  the 
posterior  end  of  a  strong  pair  of  short  muscle  bounding  ridges  which  project  anteriorly 
as  much  as  dorsally.  Short  median  septum  starting  close  to  the  apex  and  dividing 
the  muscle  bounding  ridges  before  stopping  abruptly,  leaving  an  inclined  slope  about 
2  mm.  long  at  its  anterior  end,  which  merges  with  the  valve  floor.  Diductor  scars 
short,  leaving  concentric  growth  ridges.  Adductor  scars  elongate,  close  to,  and 
partially  on,  the  median  septum.  Shell  thick  posteriorly  with  prominent  pseudo- 
punctae,  thin  anteriorly,  often  with  an  interior  reflection  of  the  exterior  ornament. 

Brachial  interior.  Widely  divergent  prominent  socket  plates  which  curve  slightly 
posteriorly  at  their  lateral  extremities.  They  are  joined  medianly  to  the  bilobed 
cardinal  process  lobes,  which  are  directed  ventrally  and  slightly  posteriorly.  The 
process  lobes  are  connected  to  a  weak  shaft,  which  in  some  specimens  bifurcates 
anteriorly,  in  others  reunites  to  form  a  weak  median  septum.  Trans-muscle  septa 
variably  developed,  but  including  in  all  specimens  a  prominent  anterior  pair  stronger 
than  and  subparallel  with  the  median  septum.  These  septa  are  often  slightly  flared 
ventro-laterally.  Owing  to  the  mass  of  structures  in  the  brachial  valve  presumably 
mainly  used  for  the  support  of  adductor  muscles  (although  some  could  have  been 
rudimentary  brachiophore  supports)  the  scars  and  shape  of  the  muscles  are  not 
readily  distinguishable.     Thick  shell  with  prominent  taleolae  antero-laterally. 

Lectotype  (here  selected).  BB  31447  (Plate  1,  figs,  r,  2),  a  brachial  valve  in 
the  Davidson  collection.  One  of  a  number  of  specimens  which  made  up  the  material 
for  the  composite  figures  cited  by  Bancroft.  With  the  specimens  there  is  a  label 
in  Davidson's  handwriting  '  Strophomena  antiquata  Sow.  bed  D.  Wenlock  Shale, 
Pentland  Hills,  found  by  Mr.  Henderson.'  The  specimens  are  from  an  horizon 
now  known  to  be  of  Upper  Llandovery  age  in  the  North  Esk  Inlier  of  the  Pentland 
Hills,  Scotland  (Mitchell  &  Mykura  1962  :  12  et  seq.). 

Dimensions  (in  cm. — all  specimens  from  type  locality) 


BB  31447  Lectotype.     Brachial  valve 
B  8485        Brachial  valve 
BB  31450  Brachial  valve 
B  13614      Pedicle  valve 
BB  31448  Pedicle  valve 


0-91 

approx.  1 

0-92 

broken 

o-86 

i-39 

0  •  72 

1-29 

0-93 

approx.  1 

1 

w. 

appro: 

<.  I 

•3 

approx. 

2 

'2 

I- 

39 

approx. 

2 

'3 

STROPHOMENACEAX  BRACHIOPODS  291 

Discussion.  In  erecting  the  species,  Bancroft  (1949  :  13)  quoted  those  of  David- 
son's figures  of  Strophomena  antiquata  which  came  from  the  Pentland  Hills.  He 
also  figured  a  specimen  (Plate  1,  fig.  10)  which  is  of  part  of  a  pedicle  internal  mould. 
This  specimen  is  not  in  the  Sedgwick  Museum  and  Mr.  A.  G.  Brighton  informs  me 
[in  litt.  March,  1967)  that  its  whereabouts  are  unknown.  In  many  copies  of  Ban- 
croft's privately  published  paper,  Dr.  A.  Lamont  has  deleted  reference  to  the  figure 
in  the  species  description,  and  substituted  '  S.  cf .  penkillensis  '  as  the  caption  for 
Plate  1,  fig.  10,  and  indeed  the  specimen  is  probably  of  the  latter  species  and  may 
be  attributed  to  Katastrophomena  of  the  present  paper.  Thus,  to  stabilize  the 
identity  of  Pentlandina,  Davidson's  figures  have  been  selected  as  lectotypes  of  the 
type  species  by  Havlicek  (1968  :  75).  The  original  specimens  used  by  Davidson 
are  selected  above. 

Apart  from  the  Pentland  Hills,  the  species  has  not  been  found  in  form  identical 
with  the  type.  There  is,  however,  a  larger  form,  represented  as  yet  by  only  one 
pedicle  and  one  brachial  valve,  probably  attributable  to  P.  tartana,  found  in  one 
locality  in  the  southern  Welsh  Borderland.  This  is  at  Cullimore's  Quarry,  Charfleld 
Green,  Gloucestershire,  which  lies  in  Tortworth  Beds  of  C6  age,  part  of  the  Tortworth 
Inlier.     The  dimensions  are  as  follows  (in  cm.). 

BB  31470  Pedicle  valve 
BB  31471  Brachial  valve 

The  larger  size  may  well  be  a  phenotypic  feature,  as  there  seem  to  be  no  differences 
in  ornament,  internal  structures  or  general  proportions  between  the  Tortworth  and 
Pentland  Hills  specimens. 

Pentlandina  parva  Bancroft 
(PI.  i,  figs.  7,  8) 

f949.     Strophomena  {Pentlandina)  parva  Bancroft  :  13,  pi.  1,  fig.  9. 
1951.     Holtedahlina  parva  (Bancroft)  Williams  :  118,  pi.  7,  figs.  8-10. 

Discussion.  A  full  description  of  the  species  is  given  by  Williams  (1951),  and 
photographs  of  it  are  included  in  the  present  paper  only  for  completeness  and  for 
comparison  with  P.  tartana  and  P.  parabola.  The  species  is  rare,  it  has  so  far  been 
recorded  only  from  the  type  locality,  in  Uzmaston  Beds  (Upper  Llandovery)  of  the 
Frolic,  south-west  of  Uzmaston  Farm,  Haverfordwest,  Pembrokeshire.  This  is  the 
same  locality  (Locality  K  of  O.  T.  Jones  on  Sedgwick  Museum  labels)  from  which 
comes  Leptaena  quadrata,  dealt  with  later  in  this  paper,  and  it  is  interesting  to  note 
that  neither  species  has  been  found  except  at  the  type  locality. 

To  judge  from  the  figures,  the  species  seems  to  have  a  close  relative  in  North 
America,  the  P.  cf.  parva  of  Boucot  et  al.  (1966  :  25,  pi.  6,  figs.  16-18,  pi.  7,  figs. 
1-10).  The  American  species  is,  however,  twice  the  size  of  the  Welsh,  no  mention 
is  made  of  any  prominent  fold  and  sulcus,  and  the  arrangement  of  the  brachial 
interior  seems  closer  to  P.  parabola  than  to  P.  parva ;  all  these  features  give  a  definite 
impression  of  specific  difference. 


292  BRITISH  LOWER  SILURIAN 

Pentlandina  parabola  sp.  nov. 

(PI.  i,  figs.  9-12) 

Diagnosis.  Alate  Pentlandina  with  pronounced  fold  and  sulcus,  and  two  pairs 
of  converging  muscle  ridges  in  the  brachial  valve. 

Description.  Exterior.  Outline  semicircular  but  laterally  alate.  A  deep  ven- 
tral sulcus  and  dorsal  fold  are  present,  but  lateral  to  these,  minor  frills  are  some- 
times developed  at  the  valve  margin.  Ornament  of  parvicostellae,  with  fine  threads 
between  them,  more  prominently  developed  in  the  median  plane.  New  ribs  arise 
by  intercalation.  In  addition,  small,  even  concentric  rugae,  broken  by  the  parvi- 
costellae are  present,  forming  an  irregular  pattern.  Medium-sized  interarea,  with 
an  open  delthyrium,  bounded  laterally  by  plates  (PI.  1,  fig.  10)  and  a  vestigial 
chilidium.  Very  small  supra-apical  foramen,  atrophied  in  adult  specimens  and  not 
always  on  the  median  plane. 

Pedicle  interior.  Straight  hingeline  with  prominent  teeth  which  form  the  posterior 
end  of  a  pair  of  muscle-bounding  ridges  of  diamond  shape,  although  they  only  some- 
times meet  at  their  anterior,  enclosing  a  small  diductor  muscle  scar  which  is  weakly 
impressed.  No  information  on  the  size  and  shape  of  the  adductor  muscle  scars. 
Weak,  broad  median  septum  not  extending  anteriorlv  of  the  bounding  ridges. 
Interior  reflection  of  the  ornament  usually  seen.  Thin  shell  with  no  prominent 
taleolae  showing. 

Brachial  interior.  Divergent  socket  plates,  curving  antero-laterally.  Small  ven- 
trally  directed  cardinal  process  lobes.  Scarcely  visible  platform  and  weakly  im- 
pressed muscle  scars,  but  bounded  laterally  by  two  pairs  of  plates  starting  posteriorly 
at  the  lateral  ends  of  the  socket  plates  and  set  diagonally  so  that  they  converge 
anteriorly  (but  do  not  meet).  Each  plate  is  convex  laterally  and  set  at  an  angle 
to  the  valve  floor.  A  verv  weak  median  septum  runs  for  a  short  distance  anterior 
of  the  bounding  ridges. 

Holotype.     OUM  C13507,   a  partly  exfoliated  pedicle  valve  from  the   Purple 
Shale  (Upper  Llandovery)  of  Domas,  Shropshire.     Grid  Ref.  SJ/5936  0062. 
Dimensions  (in  cm. — all  specimens  from  Domas) 


OUM  C13507  Holotype  Pedicle  valve  . 

OUM  C13504  Paratype  Pedicle  valve 

OUM  C13505  Paratype  Brachial  valve  . 

OUM  C13509  Paratype  Brachial  valve  . 

Discussion.  Length  measurements  were  made  in  the  median  plane,  but  in  old 
individuals  there  is  some  shell  antero-laterally  to  this.  In  the  small  specimen 
(OUM  C13509)  the  fold  has  not  yet  started  to  develop. 

The  species  is  known  only  from  the  Purple  Shale  of  Shropshire  and  is  rare,  although 
it  occurs  as  2  %  of  the  population  at  the  type  locality.     Single  specimens  are  known 


1. 

w. 

0-89 

1-70 

o-68 

1-22 

roken 

1-20 

0-26 

0-51 

S  T  R  O  P  H  O  M  E  X  A  C  E  A  X   B  R  A  C  H  I  O  P  O  D  S  203 

from  two  other  localities,  Boathouse  Coppice  [Grid.  Ref.  SJ/6205  0398]  and  Devil's 
Dingle  [Grid  Ref.  SJ/6392  0547]. 

P.  parabola  differs  from  P.  parva  (PI.  1,  figs.  7,  8)  in  having  a  still  more  pronounced 
fold  and  sulcus,  less  distinct  ornament  and  in  better  differentiated  alae  (without, 
however,  being  more  transverse).  In  addition  the  strength,  proportion  and  arrange- 
ment of  the  brachial  internal  structures  are  dissimilar  in  the  two  species.  From 
P.  tartana  the  new  species  differs  in  being  more  apsacline,  more  alate,  and  in  having 
a  frilly  margin,  in  having  a  relatively  smaller  total  muscle  area  and  a  differently 
shaped  socket  arrangement  and  muscle  area  in  the  brachial  valve.  In  addition  the 
rugae  are  more  prominent,  the  interior  less  stronglv  papillose  and  the  whole  shell 
less  thick. 

Pentlandina  sp. 

(PI-  1,  %■  13) 

In  one  block  from  the  Bog  Mine,  West  Shropshire  [Grid  Ref.  SO/3510  9815], 
there  is  a  single  broken  pedicle  valve,  BB  31299,  which  may  be  referred  without 
doubt  to  Pentlandina,  on  its  general  shape,  particularly  its  prominent  sulcus,  and 
also  on  the  character  of  the  muscle  field  and  bounding  ridges.  It  seems  fairly  closely 
related  to  Pentlandina  parabola  but  its  shell  is  rather  more  irregular ;  in  addition  no 
trace  of  the  distinctive  ornament  may  be  seen,  but  this  could  possibly  be  due  to 
the  coarse  quartzite  matrix.  The  length  is  1-02  cm.  and  the  estimated  width 
approximatelv  i-6  cm. 

The  specimen  is  significant  in  being  the  earliest  representative  of  the  genus  so 
far  known,  with  an  age  of  Middle  Llandovery. 

Subfamily  FURCITELLINAE  Williams,  1965 
Genus  KATASTROPHOMENA  nov. 

Diagnosis.  Resupinate  strophomenid  with  an  ornament  of  irregular  subequal 
costellae  in  early  stocks,  but  may  be  parvicostellate  in  later  stocks.  With  dental 
plates  and  weak  trans-muscle  septa. 

Type  species.     Strophomena  antiquata  var.  woodlandensis  Reed  1917. 

Species  assigned  : 

Strophomena  antiquata  var.  woodlandensis  Reed  1917  :  902,  pi.  18,  figs.  20,  21, 
pi.  19,  figs.  1-5.     Middle  Llandovery,  Woodland  Point,  Girvan,  Ayrshire,  Scotland. 

Orthis  antiquata  J.  de  C.  Sowerby  in  Murchison  1839  :  630,  pi.  13,  fig.  13.  Wen- 
lock  Shale,  Woolhope,  Herefordshire. 

Strophomena  dura  Bancroft  1949  :  15,  pi.  I,  fig.  n,  Wenlock  Limestone. 

Strophonella  penkillensis  Reed  1917  :  900,  pi.  18,  figs.  11-13.  Upper  Llandovery, 
Bargany  Pond  Burn,  Girvan,  Ayrshire. 

Orthis  scabrosa  Davidson  1847  :  61,  pi.  13,  figs.  14,  15.  Wenlock  Limestone, 
Benthall  Edge,  Shropshire. 


294  BRITISH  LOWER  SILURIAN 

Strophomena  scotica  [and  var.  alveata]  Bancroft  1949  :  12,  pi.  i,  figs.  4-7,  non 
fig.  3.     Gasworks  Mudstone  (Lower  Llandovery)  Haverfordwest,  Pembrokeshire. 

Strophomena  woodlandensis  geniculata  [Bancroft  MS  nom  nud.]  Williams  1951  : 
117,  pi.  7,  figs.  5-7.     Upper  Llandovery  (Q),  Llandovery,  Carmarthenshire. 

Strophonella  costatula  Hall  &  Clarke  1894  :  359,  pi.  84,  figs.  15,  16.  Niagara 
Group,  Louisville,  Kentucky,  U.S.A. 

Strophomena  radiireticulata  Twenhofel  1928  :  192,  pi.  17,  figs.  1-3.  Jupiter 
Formation  (Upper  Llandovery- Wenlock),  Anticosti  Island,  Canada. 

PStrophomena  sibirica  Andreeva  in  Nikiforova  &  Andreeva  1961  :  183,  pi.  38, 
figs.  1-7. 


Species  possibly  congeneric : 

Strophomena  rugata  Lindstrom  i860  :  371,  pi.  13,  fig.  14.  Visby  Marl  (Copper 
Llandovery)  Gotland,  Sweden.     (Possibly  a  young  strophomenid) . 

Strophomena?  pectenoides  Andreeva  in  Nikiforova  &  Andreeva  1961  :  184,  pi.  39, 
figs.  1-6.     Middle  Llandovery,  Siberian  Platform,  L'.S.S.R. 

Strophomena  lindstromi  Gagel  1890  :  43,  pi.  3,  fig.  12,  Upper  Silurian,  east  Baltic. 
(May  be  a  davidsoniacean). 

Discussion.  Unfortunately  the  genera  of  Upper  Ordovician  Strophomenidae 
have  not  yet  been  evaluated  as  a  whole,  and  several  are  not  yet  comprehensively 
illustrated.  It  is  probable,  however,  that  the  fairly  compact  group  in  the  Silurian 
may  be  classified  together  in  one  genus  to  include  all  species  not  in  Pentlandina 
or  the  Leptaeninae.  All  these  species  are  resupinate  and  have  a  distinctive  irregu- 
larly costate  ornament,  at  least  in  the  early  Silurian.  In  addition,  their  internal 
characters,  although  often  dissimilar  at  the  specific  level,  clearly  indicate  their 
congeneric  nature. 

The  new  genus  here  erected  to  include  these  species  is  placed  in  the  Furcitellinae, 
mainly  owing  to  the  similarities  with  Furcitella  Cooper  1956  itself,  particularly  in 
the  ornament  and  internal  morphology,  but  the  Chazy  genus  is  biconvex  and  has 
a  large  foramen.  Whether  the  two  subfamilies  Strophomeninae  and  Furcitellinae 
really  require  separation  is  another  matter;  the  division  (Williams  1965)  seems  to 
have  been  made  mainly  on  ornamental  grounds.  In  fact,  although  they  are  here 
left  in  separate  subfamilies,  Katastrophomena  shows  much  resemblance  to  Stropho- 
mena itself  in  morphology,  especially  shape,  differing  mainly  in  the  presence  of 
dental  plates  and  in  the  ornament,  although,  as  mentioned  above,  Katastrophomena 
has  end  members  with  differentiated  parvicostellae. 

The  only  other  strophomenid  with  which  the  new  genus  may  be  compared  is 
Microtrypa  Wilson  1945,  from  the  Upper  Ordovician  of  Ontario,  Canada,  which  is 
particularly  poorly  known,  but  which  appears  to  differ  in  ornament  and  brachial 
interior. 

In  the  British  Llandovery  there  are  thus  three  species,  K.  woodlandensis  and  its 
subspecies  geniculata,  K.  scotica  and  K.  penkillensis,  each  of  which  will  now  be 
reviewed. 


STROPHOM  E  X  A  C  E  A  X  BRACHIOPODS  295 

Katastrophomena  woodlandensis  (Reed) 
(PI.  2,  figs.  1-10) 

1883.     Strophomena  anliquata  (J.  de  C.  Sowerby) ;  Davidson  :  193,  pi.  15,  figs.  12-14. 

191 7.     Strophomena  antiquala  (J.  de  C.  Sowerby)  var.  woodlandensis  Reed  :  902,  pi.  18,  rigs.  20, 

21,  pi.  19,  figs.  1-5. 
1949.     Strophomena  woodlandensis  Reed  Bancroft  :  n. 
1951.     Strophomena  aff.  -woodlandensis  Reed;  Williams  :  118,  pi.  7,  fig.  4. 

Diagnosis.  Katastrophomena  with  coarse,  irregular  costae,  variable  shape,  and 
variable  brachial  internal  characteristics. 

Description.  Exterior.  Variably  resupinate,  ranging  from  flat  to  almost 
geniculate.  Ornament  of  thick  irregular  costae  of  subequal  size.  New  costae 
arise  mainly  by  intercalation,  but  sometimes  by  branching.  Prominent  growth  lines 
seldom  developed.  Large  interarea.  Delthyrium  completely  closed  by  large 
pseudodeltidium  and  chilidium. 

Pedicle  interior.  Straight  hingeline  with  prominent  teeth  which  form  the  posterior 
end  of  a  pair  of  muscle-bounding  ridges  of  variable  shape  but  which  generally  curve 
inwards  without  meeting  anteriorly.  Median  septum  variably  developed,  on  either 
side  of  which  are  faintly  impressed  blade-like  adductor  muscle  scars  inside  the  more 
strongly  developed  diductor  scars  upon  which  may  be  seen  both  faint  concentric 
growth  lines  and  also  radiating  striae.  Fairly  thick  shell,  particularly  posteriorly, 
but  large  taleolae  not  developed. 

Brachial  interior.  Strong  pair  of  widely  divergent  socket  plates  quite  separate 
from  the  small  erect  bilobed  cardinal  process.  Very  variable  minor  platform  and 
muscle  area  structures  (compare  PI.  2,  fig.  6,  8,  9).  A  median  septum  is  usually 
present,  which  bifurcates  to  a  greater  or  lesser  degree,  trans-muscle  septa  are  occa- 
sionally present. 

Lectotype,  here  selected.  B  54490,  a  pedicle  valve,  figured  by  Reed  (1917,  pi. 
18,  fig.  21)  from  the  Middle  Llandovery  of  Woodland  Point,  Girvan,  Ayrshire, 
Scotland.     Gray  Collection. 

Dimensions  (in  cm. — all  specimens  from  Woodland  Point) 

1.  w. 

B  54490      Lectotype.     Pedicle  valve  .  2-17  3-oi 

B  73012      Brachial  valve      .  .  .  2-05  3-17 

BB  31420  Brachial  valve      .  .  .  1-54  2-25 

BB  31422  Brachial  valve      .  .  .  1-78  2-37 

Discussion.  There  is  a  large  amount  of  variability  in  Katastrophomena  wood- 
landensis, particularly  in  two  respects;  the  degree  of  valve  convexity,  and  the 
development  of  internal  brachial  structures.  Both  points  may  be  seen  in  PI.  2; 
where  figs.  5-7  show  specimens  in  which  the  valve  direction  has  changed  in  the 
median  plane  by  more  than  90  degrees,  whereas  figs.  8  and  9  show  specimens  which 
are  only  slightly  concave.  Similarly  the  contrast  in  the  brachial  interiors  between 
figs.  8  and  9  is  self-evident. 


296  BRITISH  LOWER  SILURIAN 

K.  woodlandensis  is  not  common  except  at  the  type  locality,  but  the  species  is 
also  present  in  the  Middle  Llandovery  of  the  type  area.  Two  specimens  have  been 
found,  SMA  30006  (Williams  1951  :  118,  pi.  7,  fig.  4)  and  BB  31409,  collected  by 
the  author  from  B3  mudstones  in  a  small  disused  roadside  quarry  [Grid  Ref.  SN/ 
760  309]. 

A  subspecies  has  also  been  erected,  K.  woodlandensis  geniculata  [Bancroft  1949 
nom  nud.]  Williams  (1951  :  117,  pi.  7,  figs.  5-7),  whose  type  specimens  are  refigured 
here  for  convenience  (pi.  3,  figs.  1,  2).  This  is  another  rare  form  from  Cx  beds  in 
the  Llandovery  area.  It  is  best  left  as  a  separate  subspecies,  as,  although  the 
brachial  valve  could  well  be  identified  as  K.  woodlandensis  (s.  s.),  the  form  of  the 
pedicle  muscle  field  shows  some  affinity  with  K.  scotica  described  below.  Further 
collecting  at  the  type  locality  has  failed  to  produce  more  material. 

The  chief  differences  between  K.  woodlandensis  and  K.  scotica  are  in  the  shapes 
of  the  pedicle  muscle  field  and  bounding  ridges  and  in  the  stronger  median  septum 
usually  present  in  K.  scotica.  The  chief  difference  between  these  two  species  and 
K.  penkillensis  lies  in  the  differentiated  ornament  of  the  latter. 


Katastrophomena  scotica  (Bancroft) 
(PI.  3,  figs-  3-9) 

1871.  Strophomena  antiquata  (J.  de  C.  Sowerby) ;  Davidson  pars  :  299,  pi.  44,  figs.  21,  11  only. 

1949.  Strophomena  scotica  Bancroft  :  12,  pi.  1,  figs.  4,  5,  non  fig.  3. 

1949.  Strophomena  scotica  var.  alveata  Bancroft  :  13,  pi.  1,  figs.  6,  7. 

1951.  Strophomena  scotica  Bancroft;  Williams  :  116,  pi.  7,  figs.  1-3. 

Diagnosis.  Katastrophomena  with  irregular  costae.  Pedicle  muscle  field 
diamond-shaped  posteriorly  with  bounding  ridges  drawing  out  anteriorly  to  become 
sub-parallel  in  extreme  cases. 

Description.  Exterior.  Variably  resupinate.  Ornament  of  thick,  irregular 
costae  of  subequal  size.  New  costae  arise  by  bifurcation  and  intercalation.  Promi- 
nent concentric  growth-lines  often  developed.  Large  interarea,  with  at  least  a 
small  pseudodeltidium  and  possibly  a  large,  entire  one.     Large  chilidium. 

Pedicle  interior.  Straight  hingeline  with  prominent  teeth  which  form  the  posterior 
end  of  a  pair  of  variably  developed  muscle  bounding  ridges  of  curved  to  diamond 
shape,  not  meeting  anteriorly,  but  sometimes  drawn  out  and  extending  sub-parallel 
for  a  short  distance  anteriorly.  Weak  median  septum  running  from  the  apex  to 
approximately  the  ends  of  the  muscle  bounding  ridges.  On  either  side  of  this 
septum  are  sometimes  impressed  the  pair  of  small  blade-like  adductor  muscle  scars. 
Strongly  impressed  diductor  muscle  scars  on  which  concentric  growth  lines  are 
often  seen. 

Brachial  interior.  Large  pair  of  strongly  divergent  socket  plates  on  either  side 
of,  and  distinct  from,  the  erect  bilobed  cardinal  process.  Between  the  lobes  is 
sometimes  preserved  a  small  thin  blade.  Median  septum  usually  strong,  variably 
bifurcate.  Other  structures  very  variable,  trans-muscle  septa  and  muscle  bounding 
ridges  are  sometimes  weakly  developed. 


S  T ROPHO M E  N A C E  AN  BRACHIOPOUS 


297 


Lectotype,  here  selected.  SMA  32194,  a  pedicle  internal  mould,  figured  Ban- 
croft (1949,  pi.  1,  fig.  4)  from  the  Gasworks  Mudstone  (Lower  Llandovery),  cutting 
opposite  entrance  to  gasworks,  Haverfordwest,  Pembrokeshire.     Turnbull  Collection. 

Dimensions  (in  cm. — all  specimens  from  type  locality) 

1.  w. 

SMA  32194  Lectotype,  pedicle  valve       .  2-62 


SMA  32193  Pedicle  valve 
BB  31435     Pedicle  valve 
BB  31412     Brachial  valve 
BB  31443     Brachial  valve 


2-46 

2-21 

2-35 
0-90 


approx.  3  •  7 
approx.  3-5 

3-05 

approx.  2-8 

1-23 


Discussion.  The  second  specimen  figured  by  Bancroft  has  been  chosen  as 
lectotype  because  the  first  (1949,  pi.  1,  fig.  3)  is  the  counterpart  to  the  figured  ex- 
ample of  'Strophomena  '  agrestis  from  the  Slade  Beds  (SMA  32040),  and  was  pre- 
sumably illustrated  as  S.  scotica  in  error.  Bancroft  also  erected  a  variety,  S.  scotica 
var.  alveata,  without  giving  any  differences  from  the  nominal  subspecies.  As  the 
localities  are  the  same  and  the  types  of  both  subspecies  virtually  identical  (refigured 
here  PL  3,  figs.  4-8)  and  within  the  range  of  variation  found  in  the  Gasworks  Mud- 
stone,  no  subspecies  of  K.  scotica  seem  necessary. 

Williams  (1951  :  116)  selected  a  specimen  from  the  Gasworks  Mudstone  of  the 
Frolic  section  as  type  for  the  species,  but  as  this  is  not  one  of  Bancroft's  originals 
the  selection  is  not  valid.  Williams'  specimen  (refigured  here  PI.  3,  fig.  9)  is,  how- 
ever, clearly  conspecific  with  the  type  material  from  the  entrance  to  the  gasworks 
at  Haverfordwest. 


?i87i. 
1917. 


Katastrophomena  penkillensis  (Reed) 
(PL  4,  figs.  1-6) 

Strophomena  antiquata  (J.  de  C.  Sowerby) ;  Davidson  pars  pi.  44,  fig.  5 
Strophonella  penkillensis  Reed  :  900,  pi.  18,  figs.  11-14. 


mly. 


Diagnosis.     Katastrophomena  with  differentiated  parvicostellate  ornament. 

Description.  Exterior.  Gently  resupinate.  Ornament  of  fairly  fine  differen- 
tiated parvicostellae,  the  smaller  type  mere  threads.  New  ribs  arise  by  intercala- 
tion. Large  ventral  interarea,  but  smaller  on  brachial  valve  than  other  species  of 
the  genus.  No  information  on  extent  of  pseudodeltidium.  Chilidium  present. 
Occasional  prominent  growth  lines  sometimes  seen  on  valve  exterior. 

Pedicle  interior.  Straight  hingeline  with  prominent  teeth  which  form  the  posterior 
end  of  a  pair  of  muscle  bounding  ridges  of  approximately  diamond  shape,  although 
they  do  not  meet  anteriorly.  Median  septum  running  from  the  apex  to  a  position 
level  with,  or  just  anterior  to,  the  ends  of  the  muscle  bounding  ridges.  On  either 
side  of  the  anterior  end  of  the  septum  are  a  pair  of  bladelike  adductor  scars,  which 
themselves  may  run  anterior  to  the  end  of  the  bounding  ridges.  Diductor  scars 
strongly  impressed,  sometimes  with  radiating  striae  and  concentric  growth  ridges. 
Shell  of  variable  thickness  with  large  postero-median  taleolae  not  developed. 


298  BRITISH  LOWER  SILURIAN 

Brachial  interior.  Fairly  straight  socket  plates  widely  divergent.  Normal  erect 
bilobed  cardinal  process.  Variably  developed  muscle  field  structures;  trans-muscle 
septa  sometimes  seen  (PL  4,  fig.  3,  but  not  in  PL  4,  fig.  1).  Broad,  faint  platform 
between  two  more  or  less  circular  muscle  scars  which  are  weakly  impressed. 

Lectotype  (here  selected)  B  73013,  a  brachial  internal  mould,  figured  by  Reed 
(1917,  pi.  18,  fig.  n)  from  the  Upper  Llandovery  of  Bargany  Pond  Burn,  Girvan, 
Ayrshire,  Scotland.     Gray  Collection. 

Dimensions  (in  cm. — all  specimens  from  Bargany  Pond  Burn) 

1.  w. 

B  73013      Lectotype,  brachial  valve       .  1-57  2-74 

BB  31432  Brachial  valve      .  .  .  1-15  approx.  2-3 

BB  31472  Pedicle  valve        .  .  .  1-40  2-86 

Discussion.  Despite  the  localities  being  given  by  Reed  as  (1)  Penkill  (2)  Bargany 
Pond  Burn,  all  his  figured  specimens  come  from  Bargany  Pond  Burn,  where  the 
species  is,  however,  rare.  Reed  put  the  species  into  Strophonella  as  he  described  the 
hingeline  as  '  finely  crenulated  ',  but  there  is  no  doubt  that  the  hingelines  of  the 
specimens  to  hand,  which  include  all  Reed's  syntypes,  are  smooth.  It  is  surprising 
that  Reed  made  this  mistake,  especially  as  Davidson  had  already  labelled  some  of 
the  specimens  as  S.  antiquata  (fide  Reed  1917  :  901,  no  Davidson  label  is  with  the 
type  lot  today). 

Katastrophomena  penkillensis  also  occurs  rarely  in  the  higher  Llandovery  horizons 
in  Shropshire,  for  example  a  pedicle  valve  (BB  31408)  from  the  Minsterley  Formation 
(pi.  4,  fig.  6)  and  a  brachial  valve  from  the  Purple  Shale  of  the  Onny  River,  GSM 
11693. 

Thus  so  far  the  species  seems  confined  to  the  top  half  of  the  Upper  Llandovery, 
and  to  judge  both  from  their  ornament  and  general  aspect,  it  is  probable  that  the 
Wenlock  species  of  the  genus  were  derived  direct  from  some  earlier  form  such  as 
K.  woodlandensis  rather  than  from  K.  penkillensis,  despite  its  wide  geographical 
range. 

Katastrophomena  sp. 

(PL  4,  figs.  7,  8) 

In  the  material  from  Shropshire  there  are,  in  addition  to  the  specimens  of  K. 
penkillensis  from  the  upper  beds,  two  brachiopods  referable  to  Katastrophomena 
from  the  lower  horizons  in  the  northern  Longmynd-Shelve  outcrop.  These  consist 
of  a  pedicle  valve  from  the  Venusbank  Formation  of  The  Corners,  near  Betton 
[Grid  Ref.  SJ/314  025]  and  a  brachial  valve  from  the  Bog  Ouartzite  of  Bog  Mine 
[Grid  Ref.  SO/3510  9815].  The  latter  is  poorly  preserved  (BB  31451,  PL  4,  fig.  7), 
but  clearly  shows  the  furcitellinid  bifurcation  of  the  median  septum.  The  pedicle 
valve  (BB  31407,  PL  4,  fig.  8)  is  a  large  specimen  for  the  genus  and  has  a  more 
angularly  pentagonal  muscle  field  than  the  contemporary  K.  woodlandensis  geniculata, 
perhaps  more  similar  to  some  specimens  of  K.  scotica  from  the  Lower  Llandovery; 


STROPHOMEXACEAX  BR  ACH  I  O  PO  1)  S  299 

however  the  short  median  ridge  does  not  extend  anteriorly  of  the  muscle  bounding 
ridges  as  in  the  latter  species. 

There  is  no  doubt  as  to  the  generic  identity  of  these  Shropshire  specimens,  and 
thus  they  are  provisionally  described  here  until  more  material  comes  to  light.  Both 
specimens  show  marked  differences  from  the  named  species  of  Katastrophomena. 
It  is  noteworthy  that  no  material  ascribable  to  this  genus  has  so  far  come  to  light 
in  any  part  of  the  southern  Welsh  Borderland. 


Subfamily  LEPTAENINAE  Hall  &  Clarke,  1894 

The  relationships  and  status  of  the  leptaenids  have  been  discussed  above  under 
the  heading  of  the  superfamily.  Three  genera  attributable  to  the  subfamily  have 
been  found  in  the  British  Lower  Silurian,  Leptaena,  Cvphomena  and  Mackerrovia 
gen.  nov.  A  fourth  genus,  Bellimurina,  is  represented  by  a  single  specimen.  It  is 
possible  that  some  species  here  treated  as  Leptaena  might  be  put  into  such  later 
genera  as  Bracteoleptaena  Havlicek  1963  from  the  Bohemian  Wenlock,  but  in  the 
Llandovery  the  variation  is  not  enough  to  separate  such  species  from  Leptaena 
itself. 

Genus  LEPTAENA  Dalman,  1828 

Type  species.  L.  rugosa  Dalman  1828  from  the  Upper  Ordovician  Dalmanitina 
Beds  of  Fa  dalaberg,  Vastergotland,  Sweden.  A  lectotype  was  selected  and  the 
species  figured  by  Spjeldnaes  (1957,  p.  173,  pi.  7,  figs.  1,  2,  4),  and  also  by  Williams 
(1965,  fig.  252,  figs.  5a,  5b). 

The  species  problem  in  Leptaena  seems  more  acute  than  in  most  other  genera  of 
brachiopods.  After  the  mid-Ordovician  radiation,  the  subfamily  Leptaeninae  grew 
smaller  in  terms  of  generic  numbers,  and  thus  by  Llandovery  times  the  genera  were 
reduced  to  approximately  five,  and  all  save  Leptaena  itself  are  rare  and  sporadic. 
On  the  other  hand  there  is  some  variation  in  the  form  of  Leptaena  between  nearly 
every  locality  in  which  one  finds  it.  This  bears  out  the  assertion  of  Imbrie  (1956, 
p.  219) :  '  Study  of  living  populations  has  shown  that  if  sufficiently  rigorous  methods 
are  employed,  significant  morphological  and  genetic  differences  between  two  popu- 
lations can  be  demonstrated.  Hence  species  and  subspecies  must  be  considered 
as  collective  categories,  in  the  sense  that  they  are  composed  of  local  populations  no 
two  of  which  are  identical.' 

Thus  a  very  large  number  of  specific  or  subspecific  names  could  have  been  erected 
in  the  present  work,  but  these  would  have  served  only  as  a  smokescreen  to  hide  the 
basic  truth  of  the  '  Leptaena  rhomboidalis  '  concept.  Here  was  a  successful  stock, 
essentially  unchanged  from  the  Ordovician  to  the  Carboniferous,  which  remained 
firmly  established  in  an  apparent  variety  of  ecological  niches. 

However  there  are  some  differences  upon  which  species  have  already  been  erected, 
and  in  many  of  these  cases  the  differences  do  persist  in  many  populations  from 
many  localities.  Thus  their  taxonomic  expression  is  a  positive  step  towards  their 
recognition  and  understanding. 

GEOL.   15,   6.  31 


300  BRITISH  LOWER  SILURIAN 

Havlicek  (1968)  has  referred  several  Silurian  species  to  Leptagonia  on  the  grounds 
that  their  morphogeny  has  proceeded  nearer  that  genus  than  to  typical  Leptaena. 
In  this  paper  all  the  Silurian  forms  are  retained  in  Leptaena  until  the  Devonian  and 
Carboniferous  stocks  become  better  known. 

There  now  follows  a  list  of  previously  erected  species  of  Leptaena  of  Ashgill, 
Llandovery  and  Wenlock  age,  followed  by  a  list  of  species  previously  referred  to 
the  genus,  but  which  are  here  considered  distinct  from  it. 
Ashgill  and  Silurian  species  assigned. 

Leptaena  rugosa  Dalman  1828  :  106,  pi.  1,  fig.  1.  Dalmanitina  Beds  (Ashgill)  of 
Vastergotland,  Sweden. 

Anomites  rhomboidalis  Wahlenberg  1821  :  65.     ? Wenlock  of  North  German  drift. 

Producta  depressa  J.  de  C.  Sowerby  1823  :  86,  pi.  459,  fig.  3.  Wenlock  Limestone 
of  Dudley. 

Leptaena  tenuistriata  J.  de  C.  Sowerby  in  Murchison  1830  :  636,  pi.  22,  fig.  2. 
Wenlock  of  Marloes  Bay,  Pembrokeshire. 

Leptaena  depressa  var.  vulgaris  Barrande  1848  :  84,  pi.  22,  figs.  6,  7.  Wenlock/ 
Ludlow  of  Bohemia,  Czechoslovakia. 

L.  quadrilatera  Shaler  1865  :  65.     Ellis  Bay  Formation,  Anticosti  Island,  Canada. 

L.  schmidti  Gagel  1890  :  50,  pi.  5,  fig.  28.  Lyckholm  Fm  (Upper  Ordovician) 
East  Baltic. 

L.  richmondensis  Foerste  1909  :  211,  pi.  4,  fig.  10.     Richmond  Group,  Ohio,  U.S.A. 

L.  richmondensis  var.  precursor  Foerste  1909  :  211,  pi.  4,  fig.  11.  Richmond 
Group,  Ohio,  U.S.A. 

L.  rhomboidalis  '  var.  fi  '  Reed  1917  :  872,  pi.  13,  fig.  1.  Whitehouse  Group, 
Girvan,  Ayrshire. 

L.  rhomboidalis  'var.  7  '  Reed  1917  :  872,  pi.  13,  figs.  2,  3.  Drummuck  Group, 
Girvan,  Ayrshire. 

L.  rhomboidalis  '  var.  S '  Reed  1917  :  872,  pi.  13,  fig.  4.  Mulloch  Hill  Sandstone 
(Lower  Llandovery),  Girvan,  Ayrshire.     (Here  described  as  L.  valentia  midlochensis.) 

L.  rhomboidalis  '  var.  e  '  Reed  1917  :  872,  pi.  13,  figs.  5,  6,  non  fig.  7.  WToodland 
Point  (Middle  Llandovery),  Girvan,  Ayrshire.  (Here  described  as  L.  valentia  sp. 
nov.) 

L.  rhomboidalis  var.  nana  Chernychev  1937  :  67,  pi.  2,  figs.  16-18.  Wenlock  of 
Mongolia. 

L.  zeta  Lamont  1947  :  200.     Penkill  Group  (Upper  Llandovery),  Girvan,  Ayrshire. 

L.  haverfordensis  (and  var.  contracta)  Bancroft  1949  :  6,  pi.  1,  figs.  18-24.  Gas- 
works Mudstone  (Lower  Llandovery),  Haverfordwest,  Pembrokeshire. 

L.  valida  Bancroft  1949  :  6,  pi.  1,  fig.  25.  Upper  Llandovery  (Cx),  Llandovery, 
Carmarthenshire. 

L.  urbana  Bancroft  1949  :  6,  pi.  2,  figs.  1,2.  Upper  Llandovery  (C4),  Llandovery, 
Carmarthenshire . 

L.  elongata  Bancroft  1949  :  7,  pi.  I,  figs.  26,  27.  Upper  Llandovery  (Cx),  Llan- 
dovery, Carmarthenshire. 

L.  quadrata  Bancroft  1949  :  7,  pi.  1,  figs.  28-30.  Upper  Llandovery,  The  Frolic, 
Haverfordwest,  Pembrokeshire. 


S  T  R  O  P  H  O  XI  E  X  A  C  E  A  X  B  R  A  C  H  I  O  P  O  D  S  301 

L.  ?  tennesseensis  Amsden  1949  :  54,  pi.  5,  figs.  11-15.  Brownsport  Formation 
(Wenlock /Ludlow),  U.S.A. 

L.  delicata  Amsden  1949  :  55,  pi.  5,  figs.  11-15.  Brownsport  Formation  (Wenlock/ 
Ludlow),  U.S.A. 

L.  oklahomensis  Amsden  1951  :  85,  pi.  16,  figs.  29-35.  Henryhouse  Formation 
(Wenlock/Ludlow),  U.S.A. 

L.  acuteplicata  Sokolskaya  1954  :  60,  pi.  4,  figs.  1-4.  Porkuni  Stage  (Upper 
Llandovery),  Estonia,  U.S.S.R. 

?  Productus  twamleyii  Davidson  1848  :  315,  pi.  3,  fig.  1.  Wenlock  Limestone, 
Dudley,  Worcestershire.  [The  original  specimen  is  not  now  to  be  found,  but  David- 
son (1871  :  282)  later  put  the  species  into  synonymy  with  L.  rlwmboidalis.} 

Species  excluded  from  Leptaena 

Leptaena  sinuosus  and  L.  parvula  Kindle  1915.     See  Pentlandina. 

L.  wisgoriensis  Lamont  &  Gilbert  1945.     See  Cyphomena. 

L.  centervillensis  Foerste  1923.     Brassfield  Limestone,  Ohio. 

L.  julia  (Billings  1862).     See  Cyphomena. 

L.  loveni  de  Verneuil  1848.     See  Pentlandina. 

L.  bella  Williams  1951  :  119,  pi.  7,  figs.  14,  15.  Middle  Llandovery,  Llandovery. 
Probably  an  undescribed  genus,  but  the  small  amount  of  material  at  present  avail- 
able does  not  warrant  redescription. 

Three  groups  within  Leptaena  are  recognizable  in  the  British  Llandovery. 

(a)  Large  species  with  more  or  less  oval-sided  pedicle  muscle  scars. 

(b)  Large  species  with  more  or  less  parallel-sided  pedicle  muscle  scars. 

(c)  Small  species,  which  as  at  present  known  are  a  rather  less  well-knit  group  than 

either  of  the  other  two,  and  which  will  probably  be  found  to  be  an  amal- 
gam of  further  groups  when  more  material  becomes  available. 

In  the  Lower  and  Middle  Llandovery  the  first  two  groups  of  species  seem  to  have 
been  confined  to  the  Scottish  area  and  the  Anglo- Welsh  area  respectively,  and  this 
remains  substantially  true  for  the  Upper  Llandovery,  but  in  late  Upper  Llandovery 
time  there  was  some  admixture  of  the  two  groups;  thus  rare  L.  zeta  occurs  in  the 
Purple  Shale  of  Shropshire,  and  the  leptaenid  found  in  Deerhope  Burn  has  the  oval 
muscle-scar  outline. 

The  first  two  groups  could  thus  have  shared  a  mutual  ecological  niche,  as  they 
have  not  so  far  been  found  occurring  together.  On  the  other  hand,  representatives 
of  the  third  group,  the  small  species,  do  sometimes  occur  with  members  of  one  or 
other  of  the  larger  groups,  examples  of  this  being  at  Woodland  Point,  Girvan,  where 
L.  valentia  and  L.  reedi  occur  side  by  side  in  apparent  harmony,  and  at  locality 
H-G-A  in  the  Malverns  where  L.  contermina  and  L.  ziegleri  also  occur  together. 
Thus  at  least  two  separate  ecological  niches  may  be  inferred  for  species  of  Leptaena 
in  level  bottom  communities  at  this  time,  although  for  the  most  part  only  one  or 
other  of  them  was  occupied,  and  the  genus  is  absent  from  many  localities. 

Whether  this  scheme  holds  good  for  other  parts  of  the  world  during  the  Lower 
Silurian  has  not  vet  been  established.     The  only  foreign  species  of  Leptaena  so  far 


302 


BRITISH  LOWER  SILURIAN 


described  from  beds  of  Llandovery  age,  L.  acuteplicata  Sokolskaya  1954  from 
Estonia,  has  not  yet  been  illustrated  with  interior  views,  so  that  its  relationship 
with  British  species  remains  unknown. 

Some  consideration  has  been  given  to  the  possibility  of  formal  subgeneric  recog- 
nition of  the  three  groups,  but  this  has  been  withheld  as  taxonomic  splitting  of  this 
kind  does  not  seem  justified  in  a  generic  group  whi